Okay, this is interesting. But it also seems controversial because it involves predictions made by programming. I believe it creates issues related more or less to my suggestion.
Maybe your suggestion could be something done by a new source, like this interesting link that helps you discover your rarest records. By the way i had no idea how was that made: https://elias.pschernig.com/wildflower/leastobserved.html
Your suggestion could be made using the data from the website. If I think a bit more, I can create a flowchart on how to use the current data from iNaturalist for a program that makes those predictions.
But still, you raise an interesting point about the possibility of greater interaction with the data. This contributes to citizen science, doesn’t it?