This is a good paper that makes valid points about the possibilities of migrating flexible IT workloads, however it makes classic assumptions I see in most papers that discuss this topic.