That said, every time I read some post or article about how terrible the metadata for classical music is on the various streaming services, I am reminded of when I worked at Borders in the mid-1990s. Most of our stores had a fairly enormous selection of classical music on CD, and we had exactly the same problem. Even when the classical section was shelved and tended to by staff with deep knowledge of the genre, it was nearly impossible to keep it organized in a way that made consistent sense.
Do you break it up by composer? But what about compilations? How do you handle multiple performances of the same piece? What if a customer is interested in a particular orchestra? Do you shelve all works by composer, regardless of length or style?
The questions are endless, and we had to make a call on every one at the bookstore and then rely on well-trained staff to help customers find the particular recording they wanted. It was rare that a day went by that one of didn’t have to listen to some pedantic lecture from a customer about why this recording should be over there, and that fact that we had shelved it here demonstrated what philistines we were and why chain stores were a pox upon the earth.
My point in going into this whole tale is to say that categorizing classical music is complicated, and there is no easy answer to these questions. My experience on Apple Music is that the ability to search by multiple aspects of a record—e.g., composer, title, and soloist—makes the whole thing much easier.
I feel we have been trained to expect that the algorithms will automagically find exactly what we want for us, without us having to do any work. This expectation is, of course, silly and unreasonably.