Here you go, someones (slightly horrid) basic layout of how it can work.

You do not have to run a local of Musicbrainz (I do - because I can, it removes API limits but its expensive in storage and data) just point to the public instance. Also you could do Headphones, but I moved away from that years ago and have had a much happier experience with Lidarr.

So Lidarr/NZBGet (or whatever you use) are pretty straight forward.

It gets complex with beets. Not that it is inherently complex, it just has an absolute shit ton of options. You want to start with a yaml config, and just get the feel of how it operates. There are lots of "howto's" online, but unfortunately "beets" is a way to simple search. So you need to beef it with some specifics related to ahem music.

The manual and github do have it well documented. I would suggest starting with a subset of your collection, and just tinkering, (move files from /home/a to /home/b, convert to mp3 and fix ID3). It comes together pretty quickly. But the configurables of beets is crazy (in a good way).

Other things like triggering scans from LMS etc, they are documented on their respective sites.

I'll fess up - its not immediately for the faint hearted, but its probably not that hard for most people - who actually read documents and learn.