Scientists want to be published in 'Nature' because it's good for their career (citations) and respect from peers. If there was a way to fix that incentive where it was more prestigious to do the research in the open and by collaborating (like FOSS development) that would be a big win.
The tricky part is that the current system may act as a filter where the best people will still try to hide their methods and get into the existing Journals because it's best for their career - so you'll get lower quality requests for funding.
I don't understand why Universities don't work together to push towards open access, but I guess they're interested in Journal prestige too.
As a side note there's also the issue with papers hiding critical components of their research so they can't be reproduced (so they can make companies later) - this is really not in the spirit of science.
Yes you could be right about that. It's possible that in the short term an individual funder could see a drop in the quality of applicants -- but I think this will be offset by massive increase in overall impact. I think science funders underestimate how much power they have to compel change.
I think that's kind of the goal of YC research in a narrow sense, but it'd be interesting if there was a way to align the incentives so scientists collaborated more instead of hiding research so they don't get 'scooped'. At this same time this could get rid of the academic journals that charge huge amounts of money for access to publicly funded research by selling prestige to desperate researchers (so they can get more funding).
If research was somehow more like open source development where people got credit for their contributions and hypothesis, experiments and papers were worked on in the open (all open access) maybe science and research would be better.
I think scientists want this and universities would want this - there needs to be a better way to get credibility than publishing papers in famous journals.