Rewrite Executor stress test

[mjcarroll]

There was a stress test of the executor implementation introduced in: #2187 and removed via: #2206

It wasn't conclusive if the test implementation was the issue. I believe that we need to be exercising the executors in this way, but I'm not sure if the test here was the correct way to do it (given CI issues).

This test, or a rewrite of the test should be re-added to cover these use cases.