The problem has more to do with editors. The theory is that less visits leads to less editors in the long run.

I may be wrong, but I don’t think the people that edit Wikipedia are the same people that are content with half truths from LLMs and thus no longer visiting the site. So I kinda doubt it matters much.

That’s certainly true for stack overflow, but in their case the moderators were very active in getting the negative feedback loop going.