There's another thread suggesting a tag/flair system. I think that's about as far as it should go in terms of a "nemesis" list. People's bad behavior tends to be self-revealing, and I think drawing lines in the sand goes against the nature of the site. That said I'm 100% for flagging Nazis and racists.
But if the site is self moderated, for every racist user, every user will need to block a user once, where as if the shared list was crowd sourced, the action would need to be taken once. Hubski has 10 users, scenario 1: one racist user, posts a comment, 9 people filter the user Scenario 2: 10 users, one makes a list and adds the filter, same effect if subscribed to the list. Hubski has 1 million users, 10 racist users Scenario 1, each posts a racist comment - ten million filter actions must take place to 'self moderate' the racism Scenario 2, a user takes 10 filter actions publicly, same effect as with scenario 1, for all subscribed.