I've seen this point before, and it's a reasonable one, but I think there's an important distinction: some people are interested in seeing paintings in a museum but not photos, and others might be the opposite, and this is fine because it's pretty easy to distinguish between them and people who are operating in one of the missing mediums rarely try to pretend to be producing something from the other. The consensus view on "is this a painting or a photograph" is way more uniform than "is this piece of writing from AI or not", and I think that changes things.

You could totally argue that if people can't tell the difference, it's irrational for them to care which one they get, and I don't totally disagree with that either, but it's not like personal tastes have ever really been a rational thing either. Our ability to enjoy something is the result of a bunch of signals in our brains, and it's not that crazy that adding another signal (or removing one) can change that result in a way that makes it more or less desirable to seek out. Some people might literally like a piece of writing more if they have reason to believe it's from a human than they would if they read the same exact thing but had reason to believe it's from an AI, and while I would find a study showing that as fascinating, I wouldn't see that as an argument that people like the wrong things, because "right" or "wrong" don't really seem like they apply to that sort of thing. If someone told me that knowing there's a human on the other end and that having some sort of indirect, one-way emotional connection to them is an important part of what makes them enjoy it, who am I to tell them that's wrong?