You nailed it. Ignore the publishers. Just get academic institutions to put as much weight on a publication in an open sourced journal as they do for the leading journals. The problem will then solve itself.