some lessons learned so far:
#1 Dice & pseudo-random numbers: Flat distributions of single dice are fine. Flat distributions for handfuls of dice are not - no surprise. At first I was leaning on ANYDICE to get probabilities for values of multiple dice and converting it to a linear lookup - BUT - iterative accumulation results in the same distribution. And is much easier to deal with.
#2 Consoles & Windows: I haven't programmed much at all in the last decade and in that time Windows has pushed the command line interface farther and farther away. Just getting a console to behave in a predictable manner under modern Windows is kinda a pain in the ass at this point. Further reasons to go back to BSD eventually.