FWIW, I’ve been running BB on a test system that averages about 60 spam comments a day. Most of them get caught by our content filters ~80%, the other 20% end up in moderation. Since running BB, I’ve seen about 20% get cut out right away by BB2. I see now that when I test the IPs manually against various black lists (Projecthoneypot, Spamhaus, SpamCop, etc)… these IPs are not yet identified as “comment spam sources”. But on inspection, that’s clearly what they are.