We've had a lot of discussions of Simulations since
the time of this post, the credit to Keith Devlin and
his Stanford group's wish for an Apollo Project
involving simulations. This has sparked some recent
dialog on MathFuture, among other places.
One question we should ask is about how open source to
make this project. If it's a teaching enterprise,
inviting participation across the board, then it probably
shouldn't be hidden or secret code. That's not in the
spirit of liberal arts learning.