That only works if the editorial board "flips", though - unfortunately, universities mostly aren't able to start journals with reputable editorial boards.

I'm currently leaning towards finding a way to convince large numbers of boards to flip as the most viable path to proper open access.