As a researcher, in a lot of cases, you want to be published in a reputable journal. Publishers have learned to charge for the reputation over the years. Reputation has a lot of inertia.

The PR model would have to answer the question of how reputation is built. A "star" system would likely be inadequate for a number of reasons. But the major obstacle would be in getting researchers on board with the new model.

Early adopters would have to withstand scoffs from their peers who haven't accepted the new model yet.

Even Open Access, which could be regarded as an incremental improvement, has years, if not decades, of grueling work behind it.

The network effect is very strong within the research community, which results in a situation which is very close to "all or nothing."