Check for status of upgrade/delete instead of have sleeps in E2E

[vivek-koppuru]

We should be checking for the status here on whether we can perform the next action or not, whether it is a prevalidation in the delete command or a status check in the e2e tests. That way we can remove our reliance on these arbitrary 5 min sleeps.

https://github.com/aws/eks-anywhere/blob/main/test/e2e/multicluster.go#L22