NSF BIO panels are meeting right now, or have met in the last couple of weeks, depending on your panel of choice. That means funding decisions are working their way through the system and notifications will be forthcoming in the next few weeks. Unfortunately, ~90% of are going to be disappointed in the results. Also unfortunately, I have a LOT of experience being in that 90%.
While I have no problem providing advice for proposal writers based on my experience from both being on an NSF panel and through making (and theoretically, learning from) many mistakes along my path of grant writing, one of the biggest things I have learned is how to deal with proposal reviews for an unsuccessful proposal.
The first thing I do with my reviews is print them out and read them over fully. I've talked before about the NSF rankings and what they mean. I get the whole "these reviewers don't know their ass from their elbow" thing out of my system and put the reviews away for a couple of days. Then I'll go back to them and read them again, while making a list of things I need to address and things I got right. What did the panel focus on? What were common flaws perceived by more than one reviewers? How can I fix those? Will it require data or a change in the focus of the question being asked? What proposal flaws did they not come out and say, but are between the lines? What are the strengths to build on? What did reviewers not understand?
Once I have a good list, I use that instead of the reviews as a guide. For one, it separates any emotional reaction to the reviews from the revision process, but it also gives me a handy check list and reminder of what I need to address and clarify. After that, I'll set up a phone appointment with the PO. This is critical for getting a better feel for how the panel conversation went and why certain points became focal issues. It is also essential for advocating for the changes you plan to make and getting a read on whether the PO thinks those will adequately deal with the concerns. Take notes.
I can't stress enough how important it is to bounce ideas off the PO. More than once I thought I had a new strategy figured out only to have the PO say "but that's not going to address XXX". Whereas I was initially frustrated by some discussions I had with POs because it felt very one-sided, I was ignoring the subtly of the language they have to use. Talk about your revision ideas but even though you'll be doing most of the talking, you have to listen carefully to the reaction. Remember that they know exactly what people did and did not like about your proposal and that information may not appear in your panel summary, depending on how it was written.
Funding rates are not good right now, so it's important to make every application count. Remember that if there were parts of your proposal that the reviewers didn't "get" then you need to clarify the language there and make sure it is obvious what you want the reviewers to take away. It is YOUR JOB to make sure even the most distracted reviewer walks away from your proposal convinced it can work, because you have everything to lose if they don't. Your PO can help you figure out the direction to take your research questions, but you need to package it up all purty like.