It looks like MPIO isn't configured/working optimally in your SAN ISCI network.
I suggest that you also open case with HP support team to review network configuration.
Some things to note:
* Please check if "flow control" is enabled on all ISCSI switch ports.
* If you can add a fast link between SAN switches (a stacking cable for example), then you can configure all SAN ISCSI ports in the same subnet.
This will provide a simpler configuration and also might improve MPIO (as each NIC on host can access every NIC on storage and not only subset).
* You can condiser deploying a physical Veeam server, in "off-host" mode.
* If Veeam remains as VM, you might benefit try both "on-host" or "off-host" modes to check what works better for you.
The difference is whether proxy is the Hyper-V host itself, or Veeam VM.
Again - I suggest that you also open support case at HP to recheck ISCSI SAN configuration, and see if it can be improved.