Units in the Light Cone need to be revisited

[brittonsmith]

The Light Cone tool uses unitary, code_length, and dimensionless numbers somewhat indiscriminately when calculating the depth and width of a light cone slice. Since the light cone has been used mostly with Enzo data, where unitary and code_length are the same, this has not been a problem, I am quite skeptical that this will work correctly for a simulation with a different unit system.