Then I would agree that the comma seems to be defective. You can always try your thunderbolt cables for good measure. They should operate the same, as they are the same as 3.2 gen 2x2 internally (In this use anyways).
Edit - Disregard the TB suggestion, I see that you already tried that earlier on.
I received my LX6Z-7P155-U today and it was just another PZ1P-7P155-AC (Same Seller). So I am thinking the LX6Z's are non motorized and the LX6P's are the motorized ones. The seller says both part numbers have the same function.
Interesting, that is the first one that breaks the pattern that was building. Also disappointing since I was looking forward to it. Did you buy it new?
It doesn't need to be at the end. It can be put wherever. But it is the correct size to place both effected CAN Buses in the correct range. There is a more detailed explanation further back in the thread.
You can add a touch of glue to the inside male side to keep it from pushing into the case. You can also heat the top side wire of that resistor to get the solder to flow into the hole. You may end up with a weak solder joint if its only holding the resistor in on the bottom.
I would recommend waiting for the next batch of adapters since you aren't having any major issues. You can reserve one here. https://bluepilot.dev/lightning-coupler-order-page/
I would still be curious if anyone with the non motorized version would be willing to check their IPC DTCs (using Forscan) to see if there are "Lost comms with GSM" errors logged.
I am currently leaning towards it being a comma issue. I am working on it with the team. I still ordered a motorized shifter in case it is a bad shifter. I will let you guys know once I figure it out.
When you used the other (long) cable, did you have the purple coupler in place?
If you did, try both cables plugged directly into the comma. Nothing else inline.