There's a good reason for that: when AWS has problems, we are under very heavy load because many of our customers are on AWS. The last thing we want to do at that point is an emergency flip to a secondary provider.