e2e observers are nearly ready, our first usage could be `openshift-tests run-resourcewatch` (ref https://github.com/openshift/origin/blob/e945cb88da780e21c021b6c8b430454bcfb881cf/pkg/monitor/resourcewatch/cmd/resourcewatch.go).

This command creates a git repo and records every state transition for each resource as a git commit.  This will allow us to create timelines that start after the bootstrap kube-apiserver is present that include the install timeframe.  For instance, chasing the "why is this operator progressing" we'll have the data.  Or, "why is this alert firing", we'll have the data.

I think the tasks are:

1. learn how to use the run-resourcewatch command
2. given the git repo, create a stream of monitor.Events based on the git repo
3. figure out how to wrap the run-resourcewatch to wait until a valid kubeconfig is present
4. write some kind of a post-processor that can combine the monitor.Events from the test run with those created by the run-resourcewatch that runs after the run-resourcewatch and openshift-tests are both finished.