We have had no storage issues today. I noticed that this was happening on host 13 after a VM migrated off it. Then, a VM migrated later in the day off host 10 to host 13, and the issue went away on 13 but started on 10.
I'm thinking that when a VM is migrated, it causes a query of the vmList on the receiving host but not on the sending host. Thus, vpxa on the host that is sending the VM doesn't know yet that the VM isn't running there until some other VM is moved onto it and it queries the vmList to find out.