Yes, this really seems quite simple to me. We the community get to decide what is abusive and what is supportive to the long term success of Steemit.
If we use groups/guilds to find abuse (based on the guidelines set forth) to first give notice of said guidelines, then if no change is made give a down vote and if it continues then down vote until no posts make any at all.
No one will abuse the guidelines if they make 0. The only reasons people scam it is to profit. If there is no profit then it won't happen.
Realistically it would be very easy to make a bot that spots spammy activity or "abusive" activity and give that notice automatically.
Maybe I should write a post about it, though it seems like I haven't really thought of anything new.