The surreals are actually a problem in this context. As far as I'm aware, nobody's yet come up with a satisfying integral calculus that works in the surreal num

The surreals are actually a problem in this context. As far as I'm aware, nobody's yet come up with a satisfying integral calculus that works in the surreal numbers generally. Maybe you have to leave those gaps unfilled.