It was discovered that inexperienced developers with no marketing budget, who likely turned to AI simply because of a lack of other resources, saw hardly any negative impact on sales despite the AI disclosure. These games were almost certainly going to struggle even without the use of AI.

It's a different story for the more established studios with an existing following and previous titles. Game Oracle found that the use of AI by these studios resulted in a significant 40% to 60% drop in sales.

This actually makes sense. People are willing to turn a blind eye to more and larger flaws in a game if they know it comes from a small indie developer with barely any experience, than if it's coming from a well-established studio.

The matter here is what players noticed, and avoid. I see three possibilities:

1. people checking for AI disclosures, and avoiding games with them;
2. AI output being slop and making a subpar game, something along the lines of Burton's suggestion;
3. the sort of dev who uses AI is likely to make crappy games, even without it.

#1 is irrationally brushed off by the author, but I think you should gather data before brushing off hypotheses like this. #2 and #3 are lumped together in Burton's "advice" (that boils down to "if people complain about you adding shit to your sandwich, add a bit less, also if you're pouring shit in the sandwich odds are the meat is rotten too"), but they're two different beasts — one is about the tech yielding a worse product, and the other about the developers themselves being bad.