>It apparently catches some at least, I am told.

Statistically speaking it should. Do you know if it's proportionally better than just blocking 1 in 10 would-be attendees at random?