I get the feeling that there is more to this story than they are letting on especially because the impact date range ("not updated since Dec 3rd") does not have a "start" date.
Most failures result in the loss of new files, not old files -- notably those created since the last incremental. Did they get really lucky with their backup timing, or is there some sort of hierarchical storage in play?
Also, since they were able to recover 10 of 14 affected groups, it seem more like a bad tape than a bad tape drive, in which case I would think the "November" backup would likely not be bad in the same spot and could reduce the loss window to "updates made between Nov 3 and Dec 3".
