While Nature, Science, et al like to claim that they act as an important filter for garbage science, the reality is that they themselves admitted back in 2006 that 50% of what they publish turns out to be essentially false. So they're doing a terrible job at science, but provide lovely revenues for their corporations. We really do need to forget any illusion that large corporations are part of the solution, and get on with engineering a more adequate way to provide open access with some way to filter out the very worst nonsense. It won't be easy, but trying to patch a fundamentally broken model is just a waste of time.