You appear to be disagreeing with the opposite of what I said. The question isn't "What percentage of people with a carry permit would get drunk while carrying?" but "What percentage of people who go out drinking while carrying a gun have a carry permit?"

Those two answers can be completely different. For comparison, it is extremely uncommon for a human (group A) to be in my immediate family (group B), but everyone who is in my immediate family (group B) is a human (group A).

I would expect that you're right. Someone who is likely to do the second and third likely would not have the clean record needed for the first.