I encourage you to go over and weigh in on the topic of changing funding at Joan Strassman's blog. She has been discussing the changes to NSF Bio in her last couple of posts, both WRT the preproposals and why they were needed. The discussion on the second post has since evolved into how we should change the funding structure to maximize the resource we have.
All of the traditional battle axes have been dragged out. The "we should be more like those cuddly Canadians!" and the lovable "Overhead is the problem!" arguments were quick to pop up, as always. For those of you who are bored, we've discussed those issues here and here, respectively.
I am often wary of these discussions because, as Drugmonkey often points out, commenters are always going to propose whatever option benefits them the most. That's not to say that I don't think these discussion are worthwhile, but eveyone's opinion needs to be taken with a giant rock of salt, my own included.
However, one option I don't support that does not benefit me is limiting the number of grants one PI can hold. There may be studies on the optimal number of grants per lab, I have no idea, but to arbitrarily place a limit of, say, two or four grants per PI seems to me a willing attempt to close one's eyes to the variable nature of the science enterprise. And as CPP has pointed out on occasion, limiting the number of grants past a certain threshold is punitive to such a minority of PIs as to not really make a difference in the grand scheme. If only 1% of PIs have more than 5 awards (I'm making these numbers up. UPDATE: My numbers weren't so far off!), then limiting awards to 5 is putting lipstick on a pig, but it won't do much for funding levels for the other 99%.
So, readers, feel free to leave your thoughts here or join the discussion on the Strassman blog.