That's the problem, organizations don't want do do the work so they outsource it. But the journal market is not free, it's dominated by incumbents who earned the position centuries ago. As the experience with open access journals shows so far, it's near impossible to get scientists to volunteer their free labor to a new journal. Regulatory action (like requiring open access for all govt-funded science) should be taken here.
But the publishers outsource the same work to referees. Why not cut out the middle man and have the funding agencies orchestrate the peer-review? That's what they do for grants anyway.
