Research evaluation based on academic excellence or impact

On a philosophy of science mailinglist (HOPOS), a U.K. based member recently wrote: "(...) I hope this is not an abuse of the list and I appreciate that this will be of interest primarily for UK based hopoi only. There is growing concern here over the inclusion of some form of societal 'impact' factor in the next research evaluation framework (upon which a considerable portion of our funding depends). Many science folk are 'up in arms' over this but of course it bites even harder for us arts types. Emulating a recent science and engineering petition James Ladyman and I have put a petition up on the 10 Downing St. website and if you feel anywhere nearly as strongly about this as we do, please sign up."
The link is: http://petitions.number10.gov.uk/REFandimpact/ The complete petition text:

We request the reversal of the Research Councils and HEFCE policy to direct funds to projects whose outcomes are determined to have a significant ‘impact’. The arts and humanities do have such an impact, but it is typically difficult if not impossible to judge this in the short-term. Academic excellence is the best predictor of impact in the longer term, and it is on academic excellence alone that research should be judged. ‘Users’ who are not academic experts are not fit to judge the academic excellence of research any more than employers are fit to mark student essays. The UK is renowned for its creative industries. But the roots of creativity in the intellectual life of the nation need sustained support and evaluations based on short-term impact will lead to less impact in the long-term. We also request the abandonment of plans to merge subject panels based on spurious claims of disciplinary and methodological similarities. Merging panels in most cases would undermine both methodological integrity and disciplinary identities and undermine the world class research that the UK currently produces.

A French reader suggested an international reaction against this kind of research evaluation, and the petition was also mentioned at a french site "Sauvons la recherche":
http://www.sauvonslarecherche.fr/spip.php?article2916