Because the word 'flag' has a strong connotation of 'calling to the attention of others/administrators', it was always my assumption/hope that flagging things meant some human-with-authority was likely to see it, and be able to take conscious action using their privileges.

So, I 'flag' articles that are perfectly on-topic, but have abusive against-the-guidelines headlines. (I may have even, at one point or another, upvoted and flagged an article at the same time.)

I have some concern that my strategy, if misinterpreted by the algorithms, could result in my flags being discarded as signal-free.

The rise of automated moderation is giving a different meaning to the word 'flag' -- indeed closer to 'downvote'.