I would think you would see more diversity in color choice if larger numbers per color directly correlated with lesser benefits, and vice versa. For instance, a simple function such as B=0.01(√x)/x (where 'B' is the bonus for a color and 'x' is the number of nations using said color) would have enormous benefits for a color with only a few members, whie also gradually reducing the benefits for colors with many members (without eliminating the benefit entirely). The more popular color would still have an advantage, but on a more even playing field. Of course, maybe your formula does this and I'm just not seeing it. I think it would be helpful if we could see the results of the equation in graph form, based on a limited sample size...say, for a game with 400 players, just to simulate the effects.