Why are you running different version? Scenario 1 - a 3-node high scale managing 5000 APs seems like overkill to me. Why not use a single cluster with 2 nodes running the same
Scenario 2 - I don’t understand why you would want to config this way.
i am trying to see what is the best design as we moved to AWS with 5000 AP controller, so i want use the decommission Controllers to be a secondary backup in case the AWS went down for any reason, and those decommission Controllers are set in 2 ESX with limited resources so i can use only up to 2500 AP's, hope this is clear.
for scenario 1 or 2 all firmware version will be upgraded to 5.1.x,
Current status we have defendant stand alone vSZ-H as old firmware are managing old discontinued AP's for none upgraded customers.