I disagree. The truth is, some of my most used apps are the least important and/or cheapest. This is why, if we went with a pay by usage app store plan, the true winners would be the addictive, least useful apps.
I use the Spotify app 24/7, yet I use Editorial and Pythonista (both costing around $7), an app I value to a much greater degree, sometimes just -8h daily. And the truth is, while OmniOutliner is so much usefuler than say Pinterest, people spend a lot more time on Pinterest.
Don't get me wrong, I do see why the losers are probably the 99¢ apps, after all, you probably do not use those on a concurrent basis. Yet, if we go by the hourly usage, people use social networks like Pinterest and Facebook, games like Temple Run, and music services, to a much greater degree overall.
Sure but that's easily fixed by stipulating that ad supported services are not eligible for the payout.
Aha, but of course, these services and time wasters will just adapt to have no ads. Essentially, I fear that all apps will take the approach of being excellent time wasters or services (because that's what they're being paid for), and killing all the truly great apps that that have around 20 people using them, yet cost around $7. And are not as addictive as these time wasters. Some apps try to minimize the time spent in their apps (a losing strategy in this new economy) because people pay them for a fast, frictionless service (think Drafts), but put less time into them than a browser or addictive game.
But... to address your main point - yes absolutely, I'd happily give up a quarter of revenue to move to such a scheme, where my recurring revenue would be much higher because my apps are actually used by people and have a lower barrier for new users. It would be a mutually beneficial chance for developers, apple, and consumers.
However - I do grant that developers are more likely to be able to finance this transition than musicians.
Since nobody is paying for the music during the trial, and the intended result is higher revenues for the artists (and for Apple, but notably proportionally more for the artists), it is flatly disingenuous for the artists to position Apple as getting all the benefit.
What if Apple offered the artists an advance covering the 3 months based on current iTunes revenues, and then recouped this over three years by taking a higher percentage for that period from the artists who opted for it? That would solve the problem of the 3 month gap.