I think it's difficult to refute that if the funding was public, the results should be public. But the scientists have to publish in big journals to win public

I think it's difficult to refute that if the funding was public, the results should be public. But the scientists have to publish in big journals to win public funding grants, and the big journals aren't motivated to go open access and surrender their cash cow. So you're not likely to budge either the scientists or the journals by arguing about what's ethical. It seems to me like the best approach is to change the way public grants are awarded. If grants become conditional on you ONLY publishing in open access journals, well you don't have much choice do you? Ultimately this whole game was only ever about attracting the money you need to do your job. Pretty soon the expensive publishers stop getting submissions because they're all diverted to open-access journals. Of course it would never be so easy in reality. There is a pretty entrenched chicken-and-egg situation with science publications, and it will be unavoidably messy to break it.