That's what I mean by continuous automation, using rules (constraints) to dictate how the library is managed (in your case, what beets does, e.g. auto download art of the given size and format, constrain the genre etc etc).
Nice approach. I'm the developer of https://www.blisshq.com/ and other projects. The key insight there is that managing large collections is easier with rules applied continuously, rather than thoughts-in-the-moment "I'll run a batch command with a tagger".
You could have the rules encoded in your config (or the DB) and apply those. For example, someone may want a "portable mirror" of their gold library, but that means transcoding AND chopping the embedded album art to size (or maybe their player has max resolutions).
I guess having the DB means you could potentially version control the metadata changes?
At the back end of the 90s I did some work with self-organizing maps - collapsing multi dimensional space into a 2D map. I think it's interesting as a data visualisation approach. This looks similar but I'm totally out of date with where SOM went.
When you say 128d space, what are the data that are represented, exactly?
Not only do A&A offer good speeds, their customer service is exceptional. There are chat rooms on which engineers and enthusiasts reside and if you phone them you get an engineer.