Launching a 51% attack against PoS during the night when people are asleep?

[Bizmark13]

I would think that Bitcoin and cryptocurrencies in general are the most popular in developed countries where most of the population has access to electricity, high-speed Internet, and modern computer technologies. This means that most of the network hashrate will be concentrated in the United States with the remaining hashpower mostly concentrated in Western Europe.

Now that the Bitcoin network is mostly controlled by professional miners, most miners with substantial hashing power would probably keep their miners running overnight in order to maximize their profits.

So my question is, how does this affect proof-of-stake? While I can see how most Bitcoin miners might keep their computers running all through the night due to the above reason, the idea behind proof-of-stake is that literally anyone can contribute to the security of the network regardless of how powerful their hardware happens to be. People can stake using their smartphones, or with netbooks running Windows XP, etc.

But unlike professional miners, I doubt many of these "casual" stakers would keep their computer(s) running overnight to stake PoS coins.

So how does this affect the network security of proof-of-stake coins like Peercoin? Could someone with less than half of the total coin supply launch a 51% attack by timing their attack at the right moment?

EDIT: Perhaps the existence of exchanges that stake their users' coins for profit mitigates the risk of such a threat somewhat. But to what extent? (Not to mention this also brings a whole host of new problems.)

[hack_]

51% implies just that, having sufficient control of overall network.

In the case of hybrid Pow/PoS chains it is a more complicated procedure since you would have to have large % of both the hash and the coin. All of which may cost more to acquire than perceived profit from an attack.

In the case of pure PoS , I have not really heard of someone attempting it, but I would wager he/she would run into the same problems.

There just isn't any real incentive to try it apart from a learning experience.