The problem is not starting new journals, the problem is having researchers submit their work there. And the problem with that is that the name/reputation of ex

The problem is not starting new journals, the problem is having researchers submit their work there. And the problem with that is that the name/reputation of existing journals are currently used as a proxy to judge the quality of applicants for e.g. tenure tracks. As long as that's not changed, researchers are incentivised to submit their work to the traditional, "reputable" journals. (Disclaimer: I'm part of a project that aims to provide an alternative measure of quality.)