Our shared AWS account hits AWS API quota pretty hard and because of that storage e2e tests flake. See bug #1704356. AWS has one quota for whole account and we found out that it's installer who eats the most of API quota, almost twice as actual OpenShift. See https://bugzilla.redhat.com/show_bug.cgi?id=1704356#c13 for exact stats. Installer should not be so heavy on AWS API usage and use longer delays and exponential backoff between retries where it makes sense.
The installer already has exponential back-offs and we are not going to take a hit on time taken to create a cluster to be nice to other e2e tests. every body should be backing off.