I agree, @lukestokes, any new solution to flagging may have unexpected consequences. I guess the topic of downvotes often gets debated on steemit (I have seen quite a few).

Right now I am doing my own thing with downvotes - mainly for spam commenting, more rarely for plagiarism.

My preferred approach is warning first. Then if the spammer does not heed it, I use a small 10% flag to get his attention, with a promise to remove the flag if he learned the lesson. 10% can go to 20% - and so on if the behaviour does not improve.

On a bad-day, I am not beyond giving an immediate 10% or 20% flag for the most blatant cases of spam (with an explanation of course).

It rarely happens that someone flags me back. Mostly I get apologies, and then I remove the small flag. Even if they flag me back, I tend not to bother with retaliation. My high steempower means they can’t harm me. I just try to focus on changing their bad habits.

Sometimes I realise I am dealing with robots. (Robots can post plagiarised internet articles, changing just enough words to fool @originalcontent and similar plagiarism detection bots). There’s not a lot of point in giving warnings to robots.