Thanks for the comment, David. NERC in the UK seems have run into similar issues in recent years. I don't think NSERC in Canada (where I work) has this problem, but that's likely because NSERC has a very different funding philosophy that holds down the number of rejections and resubmissions (I think NSF and NERC should seriously consider adopting NSERC's model, but that's another issue...).
Seems like an agency like NSF might be able to implement a PubCred-type system much more easily than, say, all the journals in a big field of science like ecology. With NSF, there's no challenge of trying to get buy-in and coordinated action from dozens of journals in order to prevent authors from avoiding the PubCred system by submitting to non-participating journals.
