a) Miners are very picky about pools, if something is wrong they switch. This is not "in theory", it happens all the time. Do you remember Ghash.io from last summer? Where are they now?
b) PoS has a huge image problem and it's unproven on a large scale.
a) There would be enough time to coordinate a proper 51% attack, with creating chaos and damage to the network while the miners finally react. The point still remains, that increased hashrate does not increase the security.
b) I prefer to see potential in things before they are proven on a larger scale. To me, it's only a matter of time when people will start to see how foolish PoW was and how much better PoS is. I buy innovations when the majority are skeptical and uncertain, and I sell when I hear bartenders and cab drivers telling me about their new amazing discoveries in the world of technology. The herd of easy fortune seekers will always get slaughtered, so it's not very practical to run together with that herd.
a) miners are like the f-ing Ghostbusters when they fear that something they are doing is harming Bitcoin.

b) Beyond it being unproven, I don't think it will provide the sort of trust needed for a decentralized currency. Ethereum will be switching to PoS in 2016/2017, so we will sort of see then. But Ethereum isn't meant to be a currency as such, so it will be interesting to see if any conclusions can be drawn.