Improve funding mechanics

[tklebel]

The mechanics of how resources are re-distributed based on funding decisions works, but is quite ad-hoc and maybe not entirely coherent:

• all groups get base funding of x (x = .01)
• all groups get resources subtracted for proposal writing (20%)
• winner groups get a discrete amount of further funding (.15)

It would likely be better to have one coherent mathematical formulation for this.

[tklebel]

Something else that I think would be meaningful to consider is the question of large vs small grants (holding total volume of funding constant, i.e. bigger selective grants vs smaller grants with lower barrier). This concept is used in Smaldino's paper on funding lotteries (https://royalsocietypublishing.org/doi/10.1098/rsos.190194), and I am sure it would also influence the efficacy of incentives in our model.