Given that your storage needed a controller swap. I'm wondering if you don't do a fail over to the second controller assuming it is active/standby. Could just be some oddities that persisted through the controller swap.
When first time the controller fails it did not fail over nicely and we lost half of our luns, however, now both controllers are fine after the replacement( or at least looks fine).
we have PSP most recent with acitive/active ...
Well, another thought is the time correlation of losing a LUN and logs on the storage array. Have you engaged the storage vendor further on this post issue?
I was not able to find any solution/ reason for this problem. so we migrated our vms on other Esxi hosts.