You can also give AI models Nobel-prize winning world literature and ask why this is bad and they will tear apart the text, without ever thinking "wait this is some of the best writing produced by man".

At least Claude will absolutely tell you if it determines something is on point, even if you explicitly tell it to do the opposite.

I'm just pointing this out because they're not quite as 2 dimensional as you are insinuating - even if they're frequently wrong and need careful prompting for decent quality

(after the initial "you're absolutely right!" And it finished "thinking" about it)

Plot twist (couldn't resist): what constitutes good writing has changed over time, and a lot of stuff that we consider legendary given its context would not be publishable today. Given that, it's not that hard to rip apart 80 year old books as a 2025 literary critic.

I bet that you can find plenty of exactly that from the human reviews of any past winner.