SOLVED - TLDR, it was the FP resistor, see below for details.

Posted in a different thread, probably better off here.
I have a very odd situation. 2006 GT, 6spd. Car dies 3-4 seconds after starting, and throws a P0222 code. Doesn't flood, and will restart eagerly multiple times in a row. Tried 4 different TBs (2 used, 2 remans) and the remans won't respond at all, while the used ones will open after start, if I go WOT, but wont hold RPM and die, if I let go. I tested every single wire on the TB and the PCM for continuity and shorts, and they all check out. The top-of-the-engine harness, is in fact new, I got it from Mazda a year ago. Car ran fine, until one day it just wouldn't start (dies right after starting). My understanding is that even with a broken TB or pedal (also swapped that back and forth with another used one) the car will idle, just poorly and respond to throttle inputs, also poorly. It's not throwing any other codes, so I have no reason to believe it's another component. Right now, I am thinking it could the be PCM itself. Using torque, I can see the requested pedal position change as I depress the throttle pedal. but no change in the TB, even after it starts (again, unless I use the older TBs which will react to WOT only).
The last thing I did on the car was take out the clutch pedal assembly to reinforce it. After putting it back in, it wouldn't start any longer. Replaced my old TB with another used TB, it started, ran for a day and stopped starting, again.

Update - I have had a TX company replace my PCM with another one, and copied the security info over to the new PCM. I have also replaced the CPS (really the e-shaft position sensor) with a brand new Mazda part.

The car still dies within seconds of starting. Some oddities are:
1) Even after the car starts, both old throttle bodies react to the accelerator pedal input as if the rpms were 500 or under (cranking mode) - I.e.: Under 50% pedal depression, the TB doesn't react, and then after 50% it will start opening. It usually revs to 4-5k as pedal input to TB response is very skewed in startup mode. I could technically keep the car alive by keeping pedal input at 50% or more, but as soon as I let go, it dies. I don't want to keep a cold engine revving at 5k rpm.
So, it's almost as if the ECU doesn't know the car has started, and to control the TB properly. But the ECU has been swapped out now, and the problem remains.

2) With the car off, and the key in the ignition position, pressing the accelerator pedal past 50% results in the old TB's opening as well. Just like they are supposed to, I suppose, I can see it in the Torque app, and I can also hear the TB motor working, and see it working if I have the TB lay on top of the intake manifold (for testing purposes). So the TB isn't broken per se.
3) The reman TB doesn't respond at all. Not when car is off and in the ignition position, nor when the car is starting. I tried another "new" from ebay TB that looked like a reman with same results. I even reset the NVRAM and the other ECU settings, and disconnected the battery for a while, as well as cleared the codes with Torque, in hopes to reset the sensor memory of the ECU to get used to the new TBs. No difference.

4) The way the car dies and the way I can restart it consistently without flooding, feels like the fuel is being cut off. I was starting to think about the fuel pump, but being able to keep her running at high rpms makes me feel that if the pump was dead, it wouldn't run at all, let alone be able to sustain 4k rpm.
Tomorrow I'm going to try replacing the MAF sensor, just to make sure it's not that. The only reason is - I unplugged it, and it does the same thing, so I figure, either the car can normally limp-idle w/out the MAF or it's somehow the MAF.
Has anyone seen this kind of failure before (both on my and the car end)? =]
Any recommendations?

UPDATE - SOLVED!!!!
After getting a new maf, testing the 3 remaining TBs wiring on the TBs themselves, throwing on a different set of coils, just to check that too, and almost going out of my mind, i went back to what's been bugging me all along - the odd way the car dies, and never floods, which seemed like a fuel-cutoff issue.
It would either be the ecu not knowing the car needs fuel to run, or the pump not getting the power to run.

I tested the pump with a battery and hooking up a hose to the oump output line - it worked, pumping fuel out. Did the same thing while cranking. Worked. Then, tested voltage at the pump plug JUST AS THE CAR WAS STARTING. 12.5 at the moment of start, then 0 right after (hence no fuel). Was gonna rig up a solenoid to acc and battery to just drive but then...

I remembered the stories about the low and high pressure modes, jumped the resistor plug - VOILA.
Not planning on getting a new one, because F that thing. Was never a fan of different voltages for the FP in the first place. Sort of pissed that there is no CEL for this, or one was not thrown.

Presumably the P0222 was showing because i usually tries to keep the car running after starting by giving some gas, and the pedal sensors would register a change but the TB wouldnt, as the car would die from fuel starvation. Total cost diagnosing by replacing parts, including the ecu and cps $450. Total cost to actually fix: $0. Lesson learned - Always check all possible signal wires, and just because a code is thrown, doesn't mean that's the actual problem (which is beyond uncool)

 

Have you tested continuity between each of the TB connector pins and the ECU?
DTC P0222

 

Yup, and thanks for that tip. Has been solved, see update above.