Many people feel that flagging is worse than it used to be, but they don't agree at all on what should or shouldn't be flagged. That makes this feedback less actionable than one might assume.

HN gets tons and tons of threads that are critical of LLMs, so it's possible that the ones you're seeing get flagged are just below median quality and/or overly repetitive of previous discussions.