You may not know much about the scientific publication business, but I'm willing to bet you know something about the music business. You probably know that the traditional music industry is having panic attacks about people downloads and piracy (just like they used to be having panic attacks about cassette tapes).
Both the music industry and scientific publishing are currently undergoing a Marxist revolution. I'm grateful to Jay Rosen, speaking on this Scence Talk podcast, which reminds us that Marx said that a revolution occurs when the means of production change hands.
Scientists and musicians face two, dramatically different problems. The first is creating something wonderful. The second is putting that in the hands of someone who cares.
The creative process is not necessarily easy, but you can often do a lot without that much investment is special equipment. Even so, as the creative process became professionalized, some aspects of the creative process were out of range of most people. While anyone might compose a song with a beat-up acoustic guitar, not everyone had could afford to have their own recording studio. Anyone can have an excellent scientific hypothesis, but not necessarily the right tech to test it.
The second one used to be really, really hard. Musicians would have to tour, but couldn't go everywhere their fans were. Recorded music required mass production, and no one person could afford to make a factory to press vinyl records or send them to radio and retail shops, so a huge middle management sprung up to connect musicians to their fans. Scientists with completed ideas would have to go to conferences. Research publications required specialized distribution, but no one researcher could afford typesetting and layout services and buy a printing press and send journal issues to libraries around the world, so a huge management sprung up to connect scientists to their peers.
Now the "means of production" (distribution, really) is totally changing hands. The middleman is facing extinction. Now, musicians can make their own music in their homes and put it directly in the hands of their audience. See this example from Brazil. Researchers are slowly, slowly, slowly waking up to the same idea. The internet means everyone has a printing press and a distribution network.
This story suggests that open access doesn't guarantee wide dissemination. I think this is historical inertia. The journals run under the old model still do a fine job, and still have prestige. So it remains to be seen how whether there will be a boody coup or a quiet revolution.