Make more use of conformal maps

[robertbrignull]

Let's say we know the point is to the left of some vertical line and there are no flake particles to the left of that point, then we can calculate the distribution of where it will hit that line and just use that rather than simulating jumps.