I've also been using this for a while now and it works great. My only complaint is that tools for cleaning up metadata are a bit clunky if you have a very large collection. You can basically go through every item one-by-one, or you can run an automatic script to check the whole collection. It would be great if you could define certain rules to require manual review, and incrementally run the tool to improve the tags whenever you have a bit of free time to clean up your collection.

But the ideal solution would be to have some way of generating fingerprints for each audiobook, and then build up a database which matches that fingerprint to the correct metadata. That way the work of organizing and tagging large collections could be crowdsourced; this is what other communities have done.

Maybe we're not too far off from AI-assisted tools that can just figure out how to properly tag a bunch of items correctly just by looking at the filename and existing metadata. Maybe even picking up on additional contextual clues by listening to a little bit of a chapter, to check if the title of the work is mentioned at the start.