Okay, to finish my sad story, when the battery gets to be more than about two years old and the lead has been flaking off the plates, it doesn't do a good job of damping the voltage spikes created by the generator. The most common thing to fail is one of the six internal diodes. AC generators put out three phase output that is rectified and turned into DC current by those diodes. Rectified three phase output voltage is real steady and pretty smooth. When one diode fails, you lose one of those phases AND you lose two thirds of the generator's maximum current capacity. You can still find the output voltage, (as measured with a digital voltmeter), to be a nice 14.0 volts, but if you looked at the actual current flow with an oscilloscope, you'd see a nice pulse of current, a nice pulse of overlapping current, then a gap of missing current. The output voltage will drop a lot but too briefly to be seen by the voltmeter.
Also, those pulses of current induce voltages into any wires running parallel to the output wire running back to the battery. If one of those is for an engine sensor, that induced voltage might only be a few tenths of a volt, but that's huge to the Engine Computer. That's where your elusive running problems come from. It's why we always start out by suggesting unplugging the generator as a test when everything else has been tried without success.
When you found the same problem with the battery charger connected, it was caused for the same reason. Its output is rectified single phase voltage. That has a peak of about 18 volts, then it goes down to 0.0 volts, then back up again. That is 18 volts of "ripple" voltage. When the generator has a bad diode, and that circuit is putting out 0.0 volts, the other two phases are still putting out something, so ripple voltage won't be quite as high. What this boils down to is if the battery can't smooth out the voltage, the engine is likely to run worse with the battery charger operating than with the defective generator operating.
You did the first part of the charging system test already by measuring battery voltage with the engine running. That must be between 13.75 and 14.75 volts. Since that is okay, it is okay to do the second half of the tests, but that requires a professional load tester. Suppose your generator is listed as a 90 amp unit, which is a fairly common size. Under a brief full load, you're either going to get very close to 90 amps, or you're going to get very close to 30 amps along with very high ripple voltage, (which most load testers measure).
To add to the misery, because the battery is so important, it is common to go through four to six replacement generators in the life of a GM vehicle. That's when people come here for help and we explain the need to replace the battery, unless it's less than about two years old, to reduce the number of repeat failures.
Even if your generator appears to be okay, a new battery should solve the running problem by holding system voltage steady. In the rare event the battery doesn't help, that's when we'll have to dig deeper to figure out what's happening. It sounds like you already covered the likely suspects with the wiring.
Wednesday, September 9th, 2015 AT 10:21 PM