This is the exact scenario I was worried about as well. Idea seems flawed security-wise; but then again in these kind of situations there is no real security.

This is the exact scenario I was worried about as well. Idea seems flawed security-wise; but then again in these kind of situations there is no real security. I was wondering though if there are some creative ways to improve this initiative and make it more secure. Anyone has anything better than filtering eventual candidates to give shelter to according to rough "trustworthiness" of their Twitter account?