So from a high level, the configuration I saw this with was a dual orch setup managing a single security group.
You bring up a good point and maybe the orch that is to be hotfixed perhaps needs to be "orchd stop"'d first. I'm still waiting for TAC to determine the cause, but I'll mention this in the case notes once they follow up.
I did see this behavior right after applying the JHF to one orchestrator and not the other. Upon this observation, I went ahead and applied the same JHF to the other thinking that the inconsistency was the cause, but to no avail.
The documentation on Maestro addresses so many different configurations that perhaps I either misunderstood or missed the correct procedure to follow.
Thanks