backup copy jobs were designed exactly for this scenario, so that you only touch your production data once but you can then have multiple backup copies in different locations. It's also the reason why BCJ support wan acceleration and primary backups do not, since the former were designed with remote locations in mind. Also, please note that backup copy is not just a raw file level copy of the backup files, but there's a read-check-write activity in it, which means data is checked when from backup is copied into the backup copy. Any corruption happening in the backup file will be catched by backup copy. In regards to corruption in the guest os instead, those would be copied in both backups anyway, but even your solution would not help. What helps here is versioning (as long as you can revert back in time before the corruption happened you are safe) and surebackup.
EMEA Cloud Architect @ Veeam Software
Veeam VMCE #1