IDK about this implementation ...
OVER-Denouncing ought to be tracked, too, for a user's trustworthiness profile.