I left a comment on this blog entry in which I dispute the interpretation offered in the blog of the REF as having an unduly narrow defintion of impact.
Interpreted in the way the author interprets it, the REF would be too difficult to respond to effectively for all but a few exceptional cases of research. But it is one thing for a definition to hamstring us, and another for us to hamstring ourselves by misinterpreting a definition.
The latter is more likely to happen when we approach impact definitions and criteria with a rule-following rather than a rule-seeking mindset.
One other point I don’t mention in my comment — the fact that the REF is essentially a process of expert (peer) review also gives universities room to maneuver.