Comment by richardfeynman

Perhaps it's worth considering an algorithmic review of flagging abuse. You can feed a table of flagged comments with the user, the comment the user flagged, and the context, as well as HN's rules, into GPT or a similar AI to get a first approximation of which users are abusing flagging, and on which topics flagging is most abused. I bet you'd find some interesting data!