arXiv.org warehouses e-prints of scholarly articles in math, comp sci, and physics (with some stats and bio tossed in for flavor). Were arXiv to expand to include all scholarly articles, that would help us information consumers.
Currently, researchers can elect to keep their articles behind paid content walls. E.g. this paper claims that females assign a negative weight to physically attractive, high socio-economic males ("too much of a good thing"). The abstract claims that this shows that the females believe that the males will follow a mating strategy instead of a parenting strategy. A possible problem lies in their methodology, in that they gave the test females "attractive, average and unattractive male faces paired with lonely-hearts advertisements implying high, medium or low socio-economic status". I don't know how they countered the veracity problem, i.e. were the attractive & high-income ad true, the guy probably either has something else that makes him unattractive, or is too picky; either way, she won't want/get him, so she'll discount the bachelor.
Since I can't read the article, I can't critique it. By gating materials, authors essentially close-source their work to a degree. I guess they or their institutions get paid somehow in exchange for keeping their papers closed.