In (well, after) attacks like this, and really any other massive DDOS, shouldn't it be possible to identify potential botnets and try to take them out (notify their owners that they're being used, notify their hosting providers, etc) so that they can't be used again in the future?