[pixelman32]

Scientific journals don't necessarily fund research, and groups who fund research don't necessarily benefit from proceeds from journal subscriptions. Funding scientific research isn't as simple as, say, funding startups. A vast majority of scientific research, especially basic research, yields no returns on investment.

The real issue people have with the way journals work is that they seemingly don't provide a great deal of value. Imagine if GitHub worked like a science journal: You would submit your code to GitHub. They might have someone do a code review or run some tests to assure a baseline level of quality. Then they would lock your repository and charge anyone who wanted to see more than just the readme file $30 - $50. All of the proceeds from that go directly to GitHub. You, the author of the repository, get nothing.

A software developer simply would not tolerate this model, but it's the norm in science.