2011 Chevrolet Equinox LS Setting Code for P2138 & P0697

Hey yall, 
So this is my first post on here, and I'm in need of some direction. 
I am the foreman of a large shop with 13 technicians and finding time to sit down and diag a vehcile like I used to is hard, but this vehicle has me scratching my head. 
It's a 2011 Chevy Equinox 2.4L and came in with this original complaint:

CEL, fuel pump code? Was driving in neighborhood and it died completely as if all power was removed. No issues prior to that. No indication of why this may have happened. Vehicle has had repairs in the past, not from here. Fuel pump was most recent 2 years ago at RRChevy.

Technician diagnoses the concern with a communication issue at the ABS module. Installs a new Module, programs it, and vehicle now starts, and runs but then sets a code for P2138 - APP Sensor 1-2 Not Plausible & P0697 - 5V Reference 3 Circuit. Tech replaced the accelerator pedal and throttle body with no change.
Issues is then brought to me, so I performed an idle relearn and go drive it, no issues, and no codes. Next day I drive it again, and about a mile into driving it goes into limp mode and I get it back to the shop. I see code for P0697 & P2138 is set again. 
I connect a lab scope to both APP 5v references at the pedal, and both APP signals and watch the 4 channels as issues happens again, but no signal change or dropout is seen during event. I add 2 more channels to pedal ground, no change. I then back probed at the PCM instead, and still nothing can be seen that would cause an APP or 5v circuit code. The vehicle will go into limp mode now instantly,  with key on engine off then press pedal about 30% or more. Data pid shows APP 1 & 2 agree = NO then car chimes and limp mode message on dash appears. 
I performed a voltage drop test on all 6 wires from the pedal to the PCM and all are perfect at 0.00v. I installed the original pedal and throttle body during these tests to verify parts. When in limp mode and pressing the pedal the throttle body will kick open a small amount that's allowed by PCM while in limp mode and during the throttle body motor kicking open, you can see a blip on all channels for APP Signal, 5v, and ground. 

At this point I am thinking I may have a ground issue to the PCM, or a faulty PCM that may be dropping out internally on the 5v refence 3 circuit where I can't see the fault. I read if an APP 5v reference drops out, it will revert to a secondary 5v reference circuit as aback up. 

The codes seem like its a pretty clear 5v reference issue causing the APP codes but I can not verify 100%. Any help would be greatly appreciated.
Thank you! 

I can add screenshots of lab scope, or send lab scope file for ATS eScope, and Snap-on or Autel Custom PID list video record if anyone would like to see some data during the event. 

Sounds like you've done a thorough job of testing for this issue. P0697 refers to the 5v Ref 3 circuit. According to what I can find , 5v 3 is for the MAP sensor and APP2.
There is a rather generic TSB concerning water intrusion that has you check for water in an unidentified connector in the instrument panel. The Mitchell diagram shows only connector x112 between the PCM and APP, but no location appears to be given. If you can locate this connector, it may be worth while to check for issues there.
Given that exercising the pedal assembly reveals the fault, I would be inclined to blame to pedal initially. Did the shop install an aftermarket pedal? I have had trouble with dorman pedals on GM cars. Maybe you can try a known good if there's other broken Chevy's at the shop?

Thanks for the reply,
So OE pedal was replaced with a standard brand part. I did all my initial testing with the standard installed. The tech kept the OE pedal so I then tested with that pedal too and same result.
After posting the thread last night I went back in with the scanner and was looking at all the data again. I can clear all codes and relearn throttle angle. And give it a small amount of pedal and it’s all good, when applying more pedal it faults. I put it to the floor and APP 1 APP 2 agree goes to no. I saved the data to review and found with pedal quickly to the floor APP 1 read 99% while APP 2 read 88%.
It doesn’t give me what APP voltage should be just what it sees and I know APP 1 or 2 (the lower voltage) is normally half of the other higher voltage sensor. I should have another equinox or similar vehicle here and I will compare APP voltages and try the pedal as well. I also want to scope the APP signals and compare the voltage at the PCM to the voltage read by the scanner on a bit deeper level.

If there was water intrusion causing a resistance issue, wouldn’t I see that on the voltage drop test from APP signal at the pedal to the PCM?

I did look for connector x112 previously but did not pull the cluster yet. I’m working on a heat case Volkswagen now and will get back to the equinox shortly and report back with my findings.

Thank you!!