You certainly have an uncommon situation. The LH2.4 ECU seems well protected against input malfunctions. It is supposed to run a rich, limp-home mixture based on rpm in the event of way out of range sensor readings- about 5mS at idle, not a flooding mix, and I haven't seen any reports (ever) of permanent damage resulting from anything short of prolonged reverse polarity (backward jumpstarting); for example, the AMM input is protected by 5.6K series at the pin. Even nearby exhaust system welding - only rumors and speculation. Maybe you've found a new way to destroy them?
Maybe you could pick up a noid light for the convenience. Should be an auto parts store item, I'd think. You'd see if the ECU did anything like open the injectors at key on. If not, then you could use the diagnostic output test before actually cranking. Careful it doesn't fill the crankcase with fuel and wash down the cylinder walls.
In any event, it sounds like you have a good plan to get to the bottom of it, and a not-too-costly source of parts. Testing the parts first in the other car is a good idea in hindsight, but I don't think anyone would have known to do that. I hope you pin down the answer for us all.
Here are some "normal" readings I made quite some time ago. Upper trace is the ignition derived "engine speed" signal you were looking at on pin 1. Lower trace at injectors. Still about 5mS, just a whole lot more squirts during cold cranking. The 951 does not use the 5th injector your 90 has.

Cranking - using light bulb for load instead of injectors (don't wanna flood or start)

Running at idle

--
Art Benstein near Baltimore
|