Thanks for the interesting points. I sure hope these sorts of things lead to new business models, and although I'm not clear yet on how they'll work, I don't believe they can't work. If the past is a guide something probably will work sooner or later - the demand for music is not likely to go away, and musicians will probably have to be paid, or at least some of them, anyway.
Regarding the impact of digitization on production, one concern I have currently is that while the good news for artists is that it makes it easier and cheaper to create product, that is also the bad news: the resulting flood of content drives the average value down.
Digital tools made audio production and audio-for-film & video production faster and cheaper. That allowed more people to enter those markets, but also drove down the rates anyone was able to charge, while driving up the amount of work they were expected to accomplish in a given period of time.
Similarly, in music for film, TV & games, the huge supply of inexpensively made master recordings by hungry musicians has been driving down the rates for licensed music to the point where (unless you're talking about music by a hit artist) it may hardly be worth the time it takes to talk about licensing a piece of music (let alone creating it), since the fee will likely be a few hundred dollars. As with indie records, it looks to me like the people making money licensing music are the aggregators of this increasingly cheap content, not (in most cases) the creators. And even with aggregators you're looking at a commodity business that requires large scale and a lot of automation in order to be very profitable.
I don't think these are problems are insoluble, but I think they're going to take some work, including a lot of experimentation.