A friend of mines 91 Grand Prix(3.1) has been running rich for some
time now. Yesterday I took out my Snap-On Scanner to see if I could
identify any problems. Everything looked OK, EXCEPT for one "strange"
thing:

The Integrator and Block Learn values were low, but the Injector Pulse
Width was not corresponding, and reading high!? For example, at one
point, with the engine idling hot in Drive(700 RPMs), the Integrator
was @ 108, the Block Learn @ 102, and the Injector Pulse was @ 2.1 ms?
Why is the ECM not commanding a lower OPEN time for the Injectors?
The O2 Sensor is working perfectly, and the ECM is responding, but
it's not changing the Pulse Width for some reason?

My guess is there's something wrong with the MAP Sensor or sensor
hose. This engine uses Speed Density to determine fuel delivery, but
I'm not sure if the MAP Sensor is the KING of the sensors, and
over-rides all others or not? Does the ECM "care more" about the MAP
reading than the O2? Can anyone answer this for me?

I know MAP Sensors CAN go bad, rarely, but when they do, they usually
read LOWER than normal, not HIGHER, I've never heard of a MAP Sensor
going bad because of HIGHER readings!?

One other thing, what is the "normal" Injector Pulse Width for a
Mult-Port Fuel Injected engine? I think it's around 1.3 ms, but not
100% sure.

Anyone have anything to add here? Should I look elsewhere, and forget
about a fauly MAP? I truly appreciate all input!