if Running on IndexV2 (w/ SnapVault only supported up to SP17 and again from FR20), you could have requested the BackupCopy per VM.
With IndexV1 you can only recopy the complete job and hope for the best....
I ran into this issue with only one out of 140VMs covered by one job lately.
As a workaround, I mounted the Datastore's Snapshot manually and used the vCenter to register the VM failed during backupCopy (only).
Then used a streaming backup to get it to tape for longterm retention.
recopy of multiple TBs of data using NBD (which to me is the 2nd best option) instead of NAS protocol to backupCopy VMs on NFS datastores didn't seem to be a good idea.
but on the other hand, I only have to backupCopy the data only once a quarter
If there is a better Solution, I'm also interrested in that one.