In my previous post I tried to make the point that the R21 mechanism probably isn't a good bet for newly minted assistant profs to use for a starter grant. I linked to this handy-dandy "Should you apply for an R21?" page from NIAID in order to point out some shortcomings of this approach.
This sparked some interesting discussion in the comments and on the twits about how R21's are perceived not just by applicants, but also the reviewers and even the NIH. Seems like there's very little consensus on how to treat this mechanism. It's certainly not always used for "exploratory/developmental research" (preliminary data not necessary!*) as outlined in the Parent Announcement.
So dear readers, in your opinion, what the heck is an R21? And perhaps more interestingly, what do you think it should be?**
** Aside from a mechanism that benefits ME! ME! ME! ME!