"Verification checks chunk hashes; it does not prove your VM boots" is the sentence every backup dashboard should carry. Eighteen months of green jobs on top of a single consumer SSD is a painful story, and a very common one.
When Ascendra Ventures built CleanSync, which keeps Google Drive in two-way sync with a server you own and keeps a copy of every Gmail message, we treated the destructive direction as the dangerous one. Every sync is rehearsed as a dry run first, and if it would delete more files than your limit (the README default is 10), it stops and shows you the list. Before the first sync it shows what is only local, only on Drive, and different on both, and waits for a typed confirmation. Anything replaced or deleted goes to a Recovery folder first. It is a far smaller system than PBS, but the idea matches your restore-test harness: do not trust the green check, prove the outcome before you need it.
For the scratch-VM restores, do you rotate which guests get tested each night, or restore every one on a fixed schedule?