If you are still running stock ecu programming, you should take a look at your mass air flow sensor. It could be failing intermittently. It can fail electronically, which means it would need to be replaced. Or you can remove it and make sure the sensor itself is nice and clean and if not then use Mass Air Flow Sensor Cleaner ONLY which can be found at your local auto parts store. Using anything else will damage the sensor. Also check for any air leaks starting from the mass air flow sensor all the way to the intake manifold. Any cracked/loose boots or couplers will give you a random P0300 code as well. Hope any of this helps.

Alright so got it in the shop this morning to check those things, MAF looks ok, not excessively dirty, need to get some cleaner for that tho.

Then I hooked up our smoke machine that also pressurizes the system, and there is a good amount of smoke coming out of the throttle body like where the cable hooks up and from behind the TPS. I don't think that's normal is it?

No that's not normal, sounds like your shaft seals are bad.

Sure as shit, I pulled the throttle body off and disassembled it and found the shaft seals were all dried and cracked. (Ive been working on cars for about 5 years and had never heard of those seals before ) But anyway I found the seals for $15 on eBay. I'm curious what peoples recommendations would be; rebuild the throttle body myself or look for a new/ used one and just slap it in? (cuz I'm kinda dreading putting it back together with all the springs and whatnot)

I don't think new ones are available anymore, but last time I saw one it was like $400+. It's not hard to do yourself, but getting the shaft seals in workout messing them up is kinda tricky.
I believe there is a member here offering the rebuild service in the freelancer section. This guy did mine for me http://throttlebodys.com

Watch the you tube video. It's ridiculously easy to replace the seals. Unless those shaft seals were replaced, every DSM will leak from there, although even with massive leaks at the TB I don't think that's your issue.
Your maf may look clean as a whistle but that doesn't mean crap (also the maf cleaner they sell is NOT for use on Karmen vortex mafs, says so right on the can) and the only thing inside the maf that would make a difference if dirty is the tiny little metal sensor on top that can easily be cleaned with a qtip. Obviously make sure the honeycombs aren't all smashed and disrupting air flow. The problem with our mafs is usually the plug, where the wires become brittle with age. Or (as I found out after a long time chasing problems) the little rubber piece inside the plug prevents a proper connection. Either way, these things can act up sporadically with age, it doesn't hurt to have a spare lying around (most of the time you can grab one of classifieds for about $20-30, maybe a little more with pigtail). I had an issue with my awd for years, chasing idle issues and must've changes out the iac and tps ten times (a disruption in air flow will trigger tps and iac codes), until I swapped intakes (and maf) and the problem went away.
Not sure if you frequent your local junkyard, but I've learned to accumulate as many "pocket parts" as possible. Although my local yard never has 4g63s I still have a box full of tps, iac, maf, etc that I pulled off galants and 3g eclipses, Montero's, etc that all share the same part numbers. Good luck!

Alright I'll start by replacing those throttle body seals, then take it for a test drive. If it still acts up I will take a look into the MAF sensor. I did notice that the rubber inside the plug made it hard to plug in so I just pulled it out, but that didn't change anything.

I had similar issues on my 93 7 bolt and the throttle seals made it run like a charm. There is a VFAQ for this I am pretty sure, if you haven't done it already you can check it out. It was pretty easy when I did mine, just make sure the spring is wound up right. I believe I made a line with white out and when you coil the spring around it will eventually line up and that's how you know you got it back exactly where it was. Two links from the VFAQ.com

http://www.ca.dsm.org/faq/tb-shaftseals.html
http://www.plymouthlaser.com/tbor.htm

Well put the throttle body seals in today and got the car all put back together and took it for a test drive and the same thing keeps happening. So that didn't help at all.

Next step I guess is to look into the MAF. My question with that is there any way to test it? Cuz when I drive with be Snap-On scan tool hooked up and watch the data while I'm driving and when it starts acting up I don't see anything too abnormal with the MAF voltage. Is there something I should be looking for or should I just throw another part at it and hope it works??

i seem to have a couple of things in common i also have a 95 shell with a 97 motor although i didnt have a newer ecu and here is what was wrong with mine when i bought it. on the 97 motor vs the 95 the CAS is on opposite side of the cam i looked into the computer and it was the original eeprom ecm so i swapped my other stock one in to find out it ran the same. at the time i wasnt running link or evoscan so the only option for me to make the car run properly with the 95 ecu on a 97 CAS was to swap the firing order for the plug wires and injectors so i did and fixed my motor swap running poorly condition. it was halirious the plug wires when i got the car were where ever they go just make the car run no specific firing order and injectors were in stock position. one issue i ran into with the swap and here shortly im going to fix i bought a black box ecu out of a 99 and plugged it in put everything in stock position and car was firing at the wrong time so wouldn't start? i wouldn't be amazed if we have something in common with 97+ ecm issues with our cars. forgot to add my check engine light was doing similar things when the car was sold to me.

Your scan tool measures MAF voltage?
If you want to watch the MAF signal, you should watch the frequency of the MAF signal. Not really the voltage.
For example, where the car goes back and forth between open loop and closed loop is probably somewhere around 500 Hz MAF frequency.
Anyway, more throttle and RPM makes more air flow makes higher MAF frequency.

I misspoke before, my scanner doesn't read MAF voltage, it shows me the MAF g/s (grams per second) so I would assume that is the frequency of the sensor.

I took one of the other techs at my work for a ride with me and had him watching the scanner data while I drove to look for any changes once the issue started happening. He said the only thing that changed was going from closed loop running normal to OL DRV (open loop?) while it was acting up. He said the MAF data looked normal, was responding and never dropped out, same with the rest of the sensor data.

With the MAF, grams per second isn't frequency but it is still good, it's what the MAF is for. If your scan tool tells you g/s rather than just frequency, it must know what make and model your car is. With the Snap-on tool, that would be the menu item called "Make Specific Enhanced" I think.

It's too bad the only code you are getting is 0300. That is such a general code, it's just a misfire code. Doesn't point to any particular component for the fault. There are so many other codes in the '95 to '98 2 liter DSM code table. Funny none of these other codes show up.

When your car goes into the bad running, do you think it is possibly only running on 2 cylinders? That would be real bad.

I would be interested in knowing what the model number is of the Snap-on tool you are using. I'd take a look at the user manual for it. Does your scanner know the make and model of your engine/ecu?

I kind of expect that if all of the components are good, and you have had the ecu checked out by ECMtuning (good move that), then you might have an intermittent connector somewhere, like on the power transistor, coil pack, or some other item. The connectors on these old cars are kind of lousy. You can get brand new connector hardware from Brad Sheridan at Sheridan Engineering,
http://connectors.sheridanengineering.com/index.htm

Does it always run bad when it is in open loop? I mean, what happens if you make it go open loop by going full throttle? Does it always run bad then?

It's entirely possible it's only running on less than 4 cylinders but by the time I pull over and try to pull the injector plugs or something it stops acting up so it doesn't really help

The scanner is a brand new Snap On Ethos Pro, you do have to enter the year, make, model, and engine to even be able to pull codes or data so it is vehicle specific.

I went through the wire harness already and repaired a bunch of wires but I suppose it is possible that one of the connector intermittently isn't making connecting once it's hot. It would be nice if that's all it was but tracking down which one will be a pain.

So if I go full throttle when it's not acting up it goes into open loop and pulls just fine; but when it is acting up and already in open loop if I go full throttle it is just missing like crazy bogging down and shaking.

If you could somehow verify that it is missing on 2 cylinders, that's a pretty sure clue that the problem has to do with the power transistor or coil pack or the connectors on those 2 items or the wiring to those 2 items. I don't know much about 2g's but I think this much is the same as on the 1g, that the coils/power transistor system is divided into 2 parts (2 coils) and each coil fires 2 cylinders.

When my car was running on 2 cylinders, it was intermittent and unpredictable like you are describing, the bogging and shaking was really severe like you are describing, and in my case it turned out to be the electrical connector on the power transistor. I was lucky in that this would sometimes happen at idle, so I could put my hand on various connectors while my son was in the car keeping it running (). When I put some sideways force on the power transistor connector it immediately cleared up the misfire. That was repeatable. So it became very clear that's what it was.

I suppose it's also possible that something inside your ecu is intermittent, and that it just happened to not act up while ECMtuning was testing it. So if you had a spare known good ecu to swap in there and try that out, that might help.

Ok so I tried bending the pins inside the connectors to make them tighter on the Power transistor, coil and mass air flow and took it for a drive.

Turns out whatever my problem is, it's electrical, and it just got 1000 times worse on that test drive. Something has to be shorting. When I was driving all the gauges started going all over the place, the fuel pump relay was clicking on and off and eventually the car just died. I got it towed back to the shop but now the headlights are stuck on, the windows won't roll up, door chime doesn't work etc. My only guess is that somewhere in the wire harness some wires are shorting together

Can you get a video of what's happening and how it's running? Or do you have ecmlink to where you can post a log?

Ok well I got the problem from the last post fixed, there was a short somewhere in the engine bay wire harness and/ or fuse box. I got a new harness of line, installed it and it fixed all the electrical issues.

But I took it for a test drive and the original problem persists. I took a quick video of it acting up, it's hard to tell, and I also pan over to my scan tool so you can look over most of the data. None of the sensors seem to stop working, and all the data looks normal, except the open loop

I did try disassembling the plugs and bending the tabs tighter on the power transfer unit, ignition coils, and the mass air flow sensor and it didn't help at all

Here's a link to the video on YouTube

Incase you missed it.

In the repair manuals it tells you how to test maft.

At freezing, ambient, and close to boiling.

Test resistance at all 3 temps

What plugs are you using? You really should run the plain old copper BPR6es NGK plugs (unless you upped the boost then run a BPR7Ees). The iridium, platinum, 4-tipped stuff just doesn't run well in our cars.

Video, good.
Seems funny that your intake air temp is only 68 deg F. Does it always read 68?
I looked at my logs to compare grams per second at rpm and throttle %, with yours, but it's not valid really because my throttle body is totally different from stock, also turbo etc.
But in the first few seconds of your video where you are cruising along with 32% TP and 2780 rpm, your scan shows 19 grams per sec, mine would be more like 52 grams per second at the same conditions.

It would be cool if somebody with 2g and stock throttle body, and properly working IAT and MAF could look at these numbers in their logs.
Where your scanner says "OL DRV" does that mean it's in open loop?

There is a lot of focus on the electrical system, but it also behaves like a fuel supply problem.

There are posts on here that talk about fuel pumps misbehaving when they get hot.

We've also had fuel pressure problems related to bad fuel relays, which acted up once the car had been running a while.

OP, test the car by letting it idle for the length of time you normally would be on the road testing, and monitor your fuel pressure during this time.

Ok well here's an update. I went ahead and ordered a remanufactured MAF, a new fuel pressure regulator and a PCV valve. Installed it all and same exact issue.

Doubleclutch I just put new NGK plain platinum plugs in so not the fancy nonsense but not the copper ones either.

We're on Boost I believe the OL DRV means open loop but it's weird cuz it'll show OL, CL, and OL DRV so not 100% sure

Dogwhistle I was told it sounded like a fuel delivery issue too so I replaced everything in the fuel system, fuel pump straight from Mitsubishi the whole assemble $$$$! fuel filter, fuel pump relays, and now the fuel pressure regulator with no change.. the only thing I find weird about my fuel system is that if I let the car sit long enough or run a long time the tank builds up a lot of pressure, like when I pull the cap it blows air out, but it's done that for years and this issue only just started so idk

Get rid of the platinums. I didn't believe it either, but the coppers just run so much better and cost less than 10$ for all 4.