Haha, yeah. I've had loads of thoughts about it too, but it seems pretty hard to solve.
PoW, as in Bitcoin, is not it. We've seen what ASICs are doing ATM, you can only get an efficient setup if you throw loads of money at it.
Cloudmining generally tends to lean to "Ponzi mining". CEX.IO was doing it the right way, but they had to quit because of profits being lower than maintenance fees. Cloud mining suffers from the same flaws, in the backend, as PoW, because that's what it is (If it isn't Ponzi mining)
Artificial PoW as in Burst or these MSCoins are far from what would be ideal. Same flaws as PoW, but are less robust, higher-level. So feel free to partake in such an experiment, but see it as that. It will needs loads of evolution before an MSCoin can become really big (and I'm not talking about a temporary PnD, or even the price).

Current PoS implementations can be better, but there's still loads of work to do. High % PoS has fierce difficulty levels. Low % PoS don't distribute that much coins but tend to give it more to low stakeholders. I can see PoS is the way to go, needs some finetuning though