To update this situation for others who may encounter the problem:
We checked the server windows application log and found ESE error 474-
Exchange database checksum error -on one page in the private store. The error
had repeated before over a few weeks, somewhat infrequently, and the timing
was no coinciding with the calendar events disappearing. The latest log event
of this type was earlier in the morning. Error predated the earliest clean
backups.
The store was dismounted, and integrity check (eseutil /g) reported db as
corrupted. Recovery (eseutil /r) came to clean shutdown. Integrity repair
(isinteg -fix -test alltests) reported no errors, checksums (eseutil /k)
then offline defrag, retested db then came up clean. Did not need to use
repair (eseutil /p). Stores remounted fine.
Microsoft KB reports event 474 is 99% probability a hardware or driver
failure, Will be watching. I do recall 3-4 months back, on a couple of
occasions, our two year old discovered the server power off and UPS power off
switches. Yeow! At the time there appeared no ill effects and the DB errors
only showed up much later, but left wondering if there was some latent error
we tripped over later.