Yes - they all seem like reasonable examples and I am sure with your help we can get them fixed. But i'm saying that might not always be the case.
There might also be cases where we can improve core setup of tests to prevent these unstable situations (e.g. disabling event listeners by default or something along those lines).