Uh, oh! You just triggered my musicologist side

.
The big difference between paying by the song and paying by the album is that there really isn't a lot of profit to the musicians for a single song. It used to be that singles were a very, very small percentage of sales. Now they're the predominant sales medium, and that hurts. Composers and musicians aren't making money off of music. Entertainers who do big shows are. When I was in college, my undergrad adviser commented that we were making ourselves obsolete-that a composer using a piece of software like Finale, with a good keyboard with sound samples and good speakers could replace almost any instrument adequately, and that eventually, this would replace most live music. And that has proven to be the case. If you want a string orchestra to back up your vocals (as wasn't uncommon in the 1950s-1970s), you hired an orchestra, who then had to be paid for each performance. Now, you hire a sound engineer to make a pre-recorded digital soundtrack that you pay for once. Music has never been a terribly lucrative career, but right now is a REALLY down time for trained musicians.
My suspicion is that there is very little music from the 2000-2010 decade that will make it into the music history books in 100+ years. Possibly none. Even most popular stage productions are retreads and revivals, not new works. The instrumental genres have also really not had much new going on. The growth has been in technology, with almost no growth musically. If you look at the top shows on Broadway over the last 10 years, most of them are shows that were written prior to that. Orchestras aren't playing new works, for the most part, either. Having said that, there are a LOT of dry decades, musically speaking.
I'm guessing we'll move away from that, though. In about 50 years, we've gone from the early use of the Moog synthesizer, which absolutely SOUNDED electronic (and which, if you listen to a lot of mid 20th century music, lets you date a song quite precisely based on the amount it's used and it's sound quality) to synthesized sounds that are really indistinguishable from acoustic sounds, to the point of being able to synthesize the human voice (and therefore, letting performers be cast based on performing skills, not singing skills). You don't have the "fat lady" singing opera or the "Face for radio" Broadway artists, who have excellent voices, but may not be traditionally beautiful.
My prediction is that the pendulum will swing the other way. 3D printing is allowing instruments to be created that were only hypothetical, and were as much a math or computer science problem as a music one in the early 1990s. What it will take is someone with the level of talent that the Beatles had who can incorporate these new sounds and get the US market to accept them, the way the Beatles did for non-Western influences in US pop music in the 1960s. And there are still people who are that talented coming up who want to innovate and experiment. They just need to be in the right place, at the right time.
What I do think we may never see again is the mega-fan response that happened for the Beatles or Elvis. I suspect the music that is actually musically significant and will end up being considered THE music of the 20th century may well be the sort of thing that has a niche interest and is played in small, isolated clubs, with a loyal fanbase, but that may never fill big arenas. It's not mega-stars that make it into the music history books, but, often, the people who influenced the mega-stars. Musically speaking, the Southern Gospel and Blues influence, which led, combined with a good performer and a charismatic personality, to the stardom of Elvis, is a lot more influential than the King himself.
The best and worst part about being a musicologist and trying to analyze the present? You'll never know if you ended up being correct or totally off base. It's entirely possible that in 500 years, Justin Bieber will end up being a significant musical influence. But I doubt it.