It seems like a web of trust approach could be useful. For instance, you decide to trust fellow users X to identify who's a creep, and hence automatically adopt their creep block list. Maybe something like this already exists? 

 Yes, but it doesn't let us select whose opinion to trust. That's a next-step, I suppose. 

 IMO it's a necessary step, because each group is uniquely positioned to identify content that is unwanted to that group.  

 Exactly. 💯