I agree more information would be nice, but it may not be possible to give quite the information you've asked for.

Logically, when doing a fixup, the DB may have issues - so some of the data you'd use to estimate may not be accurate. It'd probably be best to keep track of how much data has been processed versus how large the DB is.

This means that you'd have to spend a minute or two running before you could come out with a decent estimate. It also means that estimates would be likely to be pessimistic - whitespace isn't processed, so you're likely to complete before the estimated time.

If anyone from IBM is looking at this thinking that inaccurate messages are bad, please note that pessimistic ones aren't. We like it when things finish faster than they should. ;-)

Anyway, voted up because I like the idea - more feedback would be good.
Also, if you're processing an indirect file, could we pre-pend "(1 of nnnn)" before the filename?
So it'd read "compacting (1 of nnnn) - x.nsf (16% done - 45 minutes remaining)". There's no way that an entire indirect file could have an accurate estimate, so the estimate should remain per NSF - hence putting the (1 of nnnn) message before the filename. But it would be nice to know how far it's gotten through an indirect file...

(Or even a folder, if you're feeling really generous!)

l like the idea of getting some feedback that progress is being made, although I suspect that any kind of forecasting would just be as reliable as the Windows copy file progress bar. Updates every 5% or 100 documents would even be fine for me.