Oh it's worse than that. A lot worse than that. The power outage (and likely immediate UPS failure) hit mid-database write, so the database is in an inconsistent state and, worse, spans a f!*#ton of drives in a RAID-10 array. Thus, recovering the database in a usable state is something that a data recovery company is going to have to do, for a sizable chunk of change.
Thus it makes sense that he'd be looking into cloud hosting for the site as a far stabler platform, since the responsibility for keeping the site from falling to a power outage would fall to someone else with a lot more resources than him.
And if you read the "sequence of unfortunate events" he posted on Google Docs, he didn't have anything else to do over the last weekend anyways.
