Compared to the amount of work involved after unscheduled power loss in a data centre yes. Been there done that, and I would defiantly haul buckets of diesel for 12 hours rather than spent the next months sorting out systems that had not restarted cleanly.
I don't think so. At work we set up some servers on the west coast to take over in case our main provider on the east coast went down.
I'd say the chance of one of their people getting hurt isn't really worth anyone's uptime.
Also all the single site prep in the world doesn't help if that one site is taken out completely. Keeping multiple servers in multiple areas is a must if 100% uptime, even during events like this, is key.
If you know of some way to swap out servers with 100% uptime, I'd like to hear it. Even Stack Exchange, which had that sort of plan in place, had to go 'static' for about a half hour.