These don't quite work the same. S3's different durability SLAs are ultimately about how many copies of the data there are ... storing the data in more places m

These don't quite work the same. S3's different durability SLAs are ultimately about how many copies of the data there are ... storing the data in more places more times gives more durability, but obviously costs a bit more in real terms, and that's reflected in the pricing and you get to choose. Most of the redundancy that powers EC2s availability SLA is at the infrastructural level; power, network and so on. These resources are shared by all of EC2. But physical redundancy is really the "easy" part of an availability SLA, if it were just a matter of buying more routers and UPS's, it would be straightforward to hit any number of nines. Instead, the real work is all of the practices and culture that's been built up around managing a system at scale. The change process, the deployment process, and the root-cause-analysis process each are part of this; team members wouldn't lower their standards or depart from the process based on price, it'd be a very un-natural way to do things!