I’m not sure that “broadening the pool of investigators” means that the NIH wants to fund more people with 1 R01; for example a pool in which people are getting four grants each could could be broadened by giving everyone two.
Setting that aside, please help me understand your statement that “Having a single R01 is a very unproductive investment of NIH funds.” Based on the graph in 2B, if the NIH gives an R01 to someone who does not have one, they are buying roughly 2.5 WRCRs. If they instead give that grant to someone who already has 3 R01s, they are instead buying what looks to me less than 2 WRCRs. So, from 2B, it seems to be me that making that first R01 is wise investment.
But let’s assume the situation is as represented in your panel 1A. If the NIH funds that 1st R01, they are (on average) getting a roughly equal return in WRCRs as if they are funding someone else’s 4th — with the added benefit that they can now create a much larger return (+3 WRCRs) by funding a 2nd. That opportunity does not exist for funding a 5th grant because there are diminishing returns.