[rland]

I wonder if this could be solved by a sort of collective contract:

A university (and its faculty) sign a contract to publish only in open-access journals. But the contract doesn't actually go into effect until it has signatories of n% of top universities (ivies, UCs, etc).

Therefore, the awful transition (of an individual university being disadvantaged) can be alleviated. Once the contract goes into effect, the pressure is directed towards the other party: those who don't publish open access will need to justify it.