We had issues with windows and ESX hosts not doing Round Robin load balancing. When multiple hosts picked the same 3PAR port as their "primary active" path instead of properly balancing across all 4 paths, we would encounter performance bottlenecks at the congested port.
We deployed SAN Screen, which is now Netapp OnCommand Insight. It polls perf and alias data from the SAN switches, as well as ESX and the 3PAR. One of the reports it provides is a balance index, that compares the amount of traffic between the host's HBA ports and alarms when its not 50% - 50%, which happens just about every time they build a new Windows server by hand
SAN Screen also does a good job of mapping a 3PAR lun to an ESX datastore, and down to the VMDKs on that datastore. So when I have a hot LUN I want to learn more about, I can track it down to the guest VM.