Is it really likely they would be able to get 200,000 computers (presumably using 99% of CPU) actually working (presumably undetected) for that to work?
See the historical list of botnets on wikipedia. Tens of millions of computers were compromised. And don't forget that the coins mined are literally free for the operator (besides the legal risk). If they are successful, even $20/day is an extra $600/month for doing nothing.
Does it matter? As long as the blocks are not empty, then the botnets just add to the Bitcoin network. Shows how powerful the model is. New headline: Botnets enslaved by Bitcoin!
You're right, it doesn't matter from the network's perspective. I was just dispelling the idea that it's not worth CPU mining with your botnet.
