I disagree with most of what you say here for reasons already spelled out in the comments here.
But I agree with this, for very different reasons than yours:
> The conventional way for artists to make money should come from performance.
I've played violin professionally from the time I was 8 to present (36). I've only once participated in a professional recording for a couple of reasons.
1. Putting out a recording is even more difficult in the classical world than in the pop/rock/hiphop/whatever-is-accessible-right-now world. The market for classical violin soloist recordings is big enough to support about 15 or so string players at any given moment. So you have to be that good, and have those connections, and everything that goes into that. It's a very tiny world, and if your parents didn't live in the right music-centers of the world when you were a chile and couldn't afford 100k+ for an instrument and $300-500/hr for the right teachers, you are never going to be in that group.
Lots of people fall into that category, and still don't even make it close to the top. Or not even as close to the top as I did coming from Podunk, TX.
You can fund a recording yourself now. Especially now that I make literally 5 times my best income from a year working as a violinist doing software development.
But how are you going to market it? Who is going to buy it? It's a vanity project, that I've never had any interest in.
I know this is stretching the limits of on-topic-ness, but bear with me because of point 2:
2. Making a recording can do some good things for you. It can make money, for one thing. But it can also do some bad things for you.
Namely, it really wrecks the economics of your competition. Since the advent of recording technology and the post-production work that goes into every recording, you end up sounding much more perfect and amazing on the recording than you do in real life.
Every recorded performer since the 70s has been using this technology, not just people who never got famous like me. Itzak Perlman is one of the greatest violinists to ever live. He's up there with Heifetz and Paganini. He made one of the definitive modern recordings of the Bach Sonatas and Partitas for solo Violin. But even he averaged 4 edits per measure in post prod for that iconic recording. In one of the hairiest parts of the d-minor partita, he had over 32 splices in one measure.
That recording isn't an accurate representation of what he does live, which is--or was (he's getting fairly old now and doesn't sound so great live, but whatever)--amazing.
Making recordings like that leave you competing on quality with yourself. A competition you will always lose in a live performance.
In addition, I think that making recordings the way that we do today is a terrible marketing move for most artists who are trying to make it that way. You are taking demand away from yourself by providing a cheaper alternative to hearing you that costs far less.
Why would you do that? To motivate people to come show up at one of your shows where you will, by definition, let them down because you are not quite as perfect and amazing in real life as you are after post-prod?
I can see why people do this in certain realms, but I think it is, in the long run, a disservice to yourself and to musical artists in general as a general way of doing things.
It would be really interesting (although, I admit this will never happen in our lifetimes) if some high-profile groups just stopped recording and were only willing to take money by performing. Tons of caveats about this idea. The most obvious of which is, "How do people get to be high profile without albums?" And I say fair enough to that. Don't really have an answer.
I think that recording technology has caught musicians just as flat-footed as the internet and eCommerce caught brick-and-mortar shops. The difference, though, I think is that musicians haven't really figured out how badly they are being marginalized and disrupted. Now we have a few winners every year who win really really big, and the distribution of profits is getting really skewed. I can only assume that creation of new hit pop songs that sell millions of albums will become algorithmic at some point.
That's to the detriment of all art.