There's a great editorial in this week's Nature issue: "Division of Labour". It's about the European Reasearch Comission (ERC) and its first round of grant review for the new "startup" grants. The main topic is transparency and the editor claims the ERC should publish which countries are getting the grants. There's one comment in there, however, that caught my attention:
To achieve such a reputation among scientists is going to be a difficult task, as many excellent scientists will not get funded. With funds for only about 200-250 grants and 9167 applications (a success-rate of under 3%), there are bound to be many excellent proposals which did not make the cut. Chances are, that many if not most of the grants which eventually get funded are truly outstanding. However, IMHO, to pick the top 2.7% out of over 9000 applications is just completely impossible and to many the process must seem like a lottery. And lotteries, obviously, do not promote excellence.
Even further to the point: I don't think a funding regime which leads to an impossible selection process is not really indicative of any excellence in a process at all.