As swvalcon said, Barry recommends adding activator to the base and intercoat. I have done this before without issue, so I do not think that is it.
The primer I used is "regular 2k" from SPI. That is what it is named. Yes, it is urethane primer.
I did not wet sand. I always dry sand primer. The primer was shot and sanded the day before. It was approx 12 hours inbetween applying the primer and the base. The last coats of primer were block sanded first with 320 and then 600. I was wondering myself if I missed a spot sanding when I first seen it flake off, but I didn't miss the entire door.
After sanding, I clean with waterborne wax and grease remover and a white, lint free cloth until no more dirt shows up. The towel comes out clean. I give the waterborne about 30 minutes to dry with the booth fans on.
I didn't measure metal temp sadly enough, but I did monitor air temp. If anything, the metal was too cold not too hot.
The hood doesn't have any issues that I am aware of, just the doors currently (I have not tried to peel the paint on the rest of the truck, the doors started peeling on their own while color sanding). I added a picture of the hood basically to show what it looked like before sanding and doesn't have any real relevance to the discussion.
It is quite possible that I sanded the doors first. I don't remember, but it is likely. I waited about 1 hour after applying the primer to sand (SPI says 20 minutes). I wouldn't think that SPI primer has this "healing" property if it says to sand in 20 with no recoat window. Anyone know for sure?
I don't know of anything I did different between panels or for any other vehicle I have done. I thought that the bit of grey and sand scratches on the peeled paint might be a clue to someone in the field as to what went wrong. I am fishing for ideas on the cause so I don't make the same mistake twice.