Good point that there can be an aggr snapshot that could get the data back. In a similar case we had an aggr snap at a customer and were lucky to have enough disk to create a new aggr... then ran "aggr copy -s" to copy the snapshot version of the aggregate to the new aggregate... then after were able to snapmirror the volume back from the temporary aggregate used for restore. If only "snap restore -A" had a "-t file" option... a flexvol in an aggr is like a file in a flexvol so would be really nice to have but no option...but aggr copy can be useful to recover if there is an aggr snap.
Run "snap list -A aggrname" and see if any aggr snaps...if there is a snapshot before the volume delete, I would turn off the aggr snap scheduler and increase the aggr snap reserve just to preserve it long enough to copy so it doesn't autodelete.
