Thanks, yes, that makes sense. I checked that server, and there is indeed disk corruption logged in its System Event log, and a test run of chkdsk says it's still there.
I also tried restoring a single file from its most recent full backup, and a single email (it's an Exchange server). But no corruption of the mounted file system was logged in the backup server's event log. I'm curious to know why it logged it in the past, but not today. What would trigger it? Maybe it just needs to be mounted for a long time, and Windows will eventually check it? Maybe not a Veeam issue.