Really appreciate the thoughtful reply.
While I think you make a fair prima facie case that financial value could track merit in art, as it does so in so many other domains, I would caution that 'weirdness' in the market for art could easily distort this. ('Weirdness' being the proper technical term, if I recall correctly.)
There are lots of reasons art markets are weird, imperfect substitutes, monopolies, winner-take-all dynamics. But I think the biggest factor here is that consumption of truly great art requires enormous search costs.
You can't be certain if you'll love a work unless you experience it, and no one has time to experience more than a miniscule fraction of all the art humanity produces any given day.
The impact is that producers and distributors can successfully make works of minimally acceptable quality, but then market the hell out of them to drown out other recommendations during our search for new great works.
That's why, while yes, GoT has been great, we are coming off nearly a decade dominated by reality TV with no rewatchability. Commercial art has demonstrated the capacity for perfectly fine returns with either model.
Meanwhile, I'm more frequently finding that some of my favorite works have been released for free online: Down and Out in the Magic Kingdom, John Dies at the End, The Martian, How We Got Here, or Harry Potter and the Methods of Rationality.
This shouldn't be too surprising, the idea that great works are possible without strong copyrights. Shakespeare wrote at a time where copyright protections did not apply to written plays, and you might find, according to critical acclaim, test of time, or market measures that Shakespeare has produced a few superlative plays.
If we look to the median, based on my admittedly informal hunch from working a few summers at bookstores, we have to start looking at things like middle of the road trashy romance novels, which I suspect will be a pretty comparable to median fan fiction (unchecked grammatical errors and all - for a particularly memorable, but not uncharacteristic example: https://www.theguardian.com/books/booksblog/2011/sep/12/shif... ).
So we could go a few rounds where you note great commercial works and terrible amateur works, and I cite the opposite, but I think we both know that could continue forever with no resolution.
What we really need is some kind of impossible study that has amateurs and professionals performing the same works, and then some independent quality rating. Oh, and we also need to compare works with and without copyright protections.
Fortunately there is a test case for this, thanks to how many different versions will emerge of various audiobooks. Buccafusco & Heald used mechanical turk in an interesting way as well, really interesting study design:
http://emilkirkegaard.dk/en/wp-content/uploads/Do-Bad-Things...
Now, studio time and famous actors can make audiobooks incredibly expensive to produce. Production values could range from "microphone in a closet" to extremely high budget.
Even so, the study found no appreciable difference in quality for professional and amateur works, or for works with or without copyright monopolies preventing competition.
Despite all the sound and fury, I suspect you and I are closer than it might appear. I agree the utilitarian works may need help with compensation (I might quibble that copyright is an imperfect vehicle there, but sure). And copyright probably led to some very good stuff that we wouldn't have otherwise.
I really just feel like there are enough counterexamples and indications like the study above to make the value of strong copyrights an open question, I was reacting far more to your apparent certainty than to the position itself.