I agree that this is very well written and is persuasive. The part that someone might disagree with most strongly is "Works are copied [...] without offering any search functionality or other utility to users, let alone criticism or commentary."

Many people do successfully use LLMs for search even though it's noisy and prone to errors.

I don't have the piece in front of me but to the extent it's not clear from JC's analysis, I think the point is not that there's no utility, it's that the utility is not connected to the original work, as Mike describes.