Thanks for your reply. At the moment there are two use cases in my mind:
First is about cleanup of untouched imported data. E.g., TIGER-data in the US. The user name and time-range of the import is known. Those query is pretty heavy, so public overpass server only works on small areas. As GeoDesk is pretty easy to setup on my local computer, my idea was to do this locally.
Second one is similar like hdyc, but maybe more individual (in regards of areas) and as well also a bit similar like Tag-info to see how popular (in question of users using it) a specific tag is in a specific region.