Regarding your last point, sometimes I will say "the stupidest thing that could possibly work". This is an idea that could work, but has serious and obvious flaws. I find that once a team gets used to it, saying out loud a stupid but plausible solution can frame the discussion away from "whats wrong with the idea" to "find something better".
