Comment 2 by Jos Gibbons :

Haven't we already achieved this with ArXiv?

Yes, for the physics and math communities, but there are great strides being made in other disciplines, such as the cell and molecular sciences, that are relegated to paid subscription. Out of bounds for most. ArXiv and PLoS are great examples of how things should be. It turns the potentially corrupt peer review process on it's head as well, which is useful. That said, there is a reasonable argument that this shouldn't be handed over to governments/academics completely. Rather, I think the corporate owners need to find better business models that enable them to profit from more efficient delivery while increasing access. It's the same problem the journalistic world is struggling with now. My $0.02