> I don’t think it’s (maintaining species ratio & density) fundamentally impossible to solve; if you have any ideas, let me know!

Achieving maximum density appears related to the NP-complete Knapsack problem ( https://en.wikipedia.org/wiki/Knapsack_problem )

However, for low-density areas, the bias should be too low to notice.

Trees grow closer to one another in dense forests - perhaps dividing the exclusion radius by the density factor would keep the bias low enough to be invisible?