waiting point on loop paths is incorrect

[Jay Herman]

When a waiting point is defined on a path that has a loop in it, i.e. the path crosses itself later, the simulation will assume that the waiting point is on the back half of the loop even if that node segment is not on the back half.

Suggested logic fix would be to check the subpaths for the node segment and assign the waiting point to the subpath it is on. If the node segment exists in both the front half and the back half of loop, assume that the waiting point should be on the back half.