The saga continues...

I turned my attention away from vacuum leaks as a source of my problems, because I have done everything possible with the fuel/air system to try and diagnose the problem, and not gotten anywhere but a little broker...

Yesterday I got down to testing the connections at the ECU according to the Bentley manual to see if some wiring flaw was preventing a sensor from being read or something. Here's what I got:

"Fuel pump power supply from main relay"
Directions say to short terminals 20 and 21 to ground with ignition on, and both fuel pumps should operate. They didn't. I did some further testing of the relay, and I got some funky results, BUT both fuel pumps operate when I start the car (I know, because I can hear them!), so I'm not sure what to make of this result.
"Throttle switch full-throttle contacts"
Continuity test from terminal 3 to ground with the throttle fully open should have continuity, but mine doesn't. Maybe that explains the 3-3-2 code I get when testing the switch via the OBD. I don't have a problem with full throttle operation though...
"Oxygen sensor wiring harness"
I'm supposed to get continuity from terminal 24 to ground with the oxygen sensor connected. Nope. OK, this one worries me, because no continuity means that there would be no or little signal from the sensor, but when the car is running, I get the voltage reading just fine - Bentley says it should be between 0.5 to 1 volt, and I get 0.7xxx volts pretty consistently, and it fluctuates up and down ±0.05 volts or so. So I guess the green wire has a break in it? But wouldn't I get a 2-1-2 code that says the signal from the oxygen sensor is missing, AND a 'Check Engine' light? I'll test it by grounding the connector and test for continuity to ground at the ECU and report back. It may be that my O² sensor has too high a resistance from connector to ground; my multimeter tests anything above ~5KΩ as open, but wouldn't that trigger a code as well? What should the resistance be for a cold oxygen sensor? Does it matter?

 

OK, no continuity there, so I can safely assume the ECU is not getting a signal from the oxygen sensor, but that doesn't square with the fact that I don't get any "oxygen sensor signal missing" codes. I'm going to have to trace down the wiring diagram and find a convenient place to jumper a wire from the O² sensor to somewhere that directly leads to ECU terminal 24.

Assuming that's the problem, that is...

 

An O2 sensor, missing a signal or missing altogether, doesn't make these cars run poorly as you describe. But may be other ECU issues as you have exhausted all other possibilities.

 

My ECU is the notorious pink label variety, so you may have a point...

 

The weakness of the "pink label" ECU is in the circuit that drives the fuel pump relay. When they fail the fuel pumps don't run leading to a no-start condition.

However, since they pink label ones tend to fail it gives you one more good reason to source a replacement "white label" ECU. Maybe it will help some of your drivability problems.

 

That makes sense, so just to be sure, I took it for a drive with the relay bypassed so both pumps are on even with the key off. No change. Unless there's a major clog in my fuel line, which I highly doubt.
I'll look into replacing the ECU anyway, because I'm sure it'll fail eventually, but for right now, I'm pretty sure that's not the issue.

 

I also just realized that when I was testing for continuity between terminal 24 at the ECU and the oxygen sensor, I had grabbed the wrong end. So I was testing for continuity between an unplugged O² sensor and ground.
Granted, the diagnostic process did say that it should read anyway, but at least I know there isn't a break in the wire between the connector and terminal 24.

Ugh.