And that's the rub, isn't it? Once a player like Nature or Elsevier can entrench itself in the process by which the other institutions function (Like being gatekeepers on the government's merit metrics for funding), they can route around the fact that nearly every working academic doesn't actually want to support their business model by effective fiat
Personally, I think it's always a mistake to build a business into government policy. It uses the government to protect the business from the free market, and the business to protect the government from public influence or public transparency