It just doesn't do anything that isn't already handled or isn't amenable to automation. My cards and cameras are recognised when they connect using USBDLM (https://www.uwe-sieber.de/usbdlm_e.html) which fires off batch routines that move the files to the right part of my file structures, add in additional files where needed for later web activity, or transfers race result spreadsheets and text files as needed. Ejects the device/card when it has verified that all files have been successfully moved. If I'm remote on the Laptop then they stay local to the Laptop, if the Laptop is on the local network then they are transferred direct to the Server.
All my images have GPS data in the EXIF direct from the cameras, so no need to handle separate GPX/KML etc. files to GeoTag. You can't rely on Google location info as you move around a site to remain consistent, it will hop around the nearest road names or not have any useful reference point other than a Town name at best. As I do a lot of watersports, I'm sometimes on a RIB in the middle of a river or estuary Google usually returns "Europe", so Reverse lookup isn't much help!
A lot of my labelling is tied up with event activities, specific race starts, turns, finishes etc. that can't be easily grouped without looking at the image and image time and my own commentary captured using a headcam and image timings for things like Podium prize givings etc.
The other major activity is labelling individuals within an image, again not something that can be done at import time, I tend to now use TagThatPhoto running on the laptop when out and about, and also on the main workstation, but this takes place in background when driving home or overnight ready for next day etc.
Would be interested how others use the features and facilities, especially now when we all have time to think about about our processes and workflows!
Below is a typical set of race Event structures I use, expanded to show all the levels.