Thank you for this valuable contribution to the community. I have struggled with the concept of neglectedness - and it predictive value for marginal returns - since it was first introduced to me. I don't buy it and your post gives a first good counter argument.
One aspect that seems to be overlooked is the effect of R&D on marginal returns. If more funds have been spent on R&D, funds spent on implementation can be allocated more efficiently. So for many cause area's you may actually see increasing marginal returns (on dollars spent on implementation) as the area matures.
For example, the amount of dollars required to save one QALY within the cause area "HIV" by providing medication today is substantially less than 10 years ago, and this amount will continue to decrease. In other words, the function of "additional QALY saved from HIV by investing one dollar of medication at time t" could be decreasing in the "amount of funds invested in implementation up to time t" (diminishing returns), but it is also increasing in the "amount of funds invested in R&D up to time t". This function is not necessarily decreasing (or increasing) in the total amount of funds invested in HIV (implementation and R&D). To make things more complex, consider the impact of the last dollar spent on HIV medication that stops the last HIV infected person from spreading and thereby effectively eradicates HIV from the earth...
[this is just an example. I do not advocate to invest in the cause area HIV]
Long story short: it all depends on the cause area and its current state. I would strongly advocate to drop the heuristic of neglectedness. It does more harm than good, and leads to a scattering of funds instead of global coordinated action on the earths most urgent issues.