Support alternative geometric definitions of the first wall

[gonuke]

Currently, the first wall is defined by evaluating a VMEC description of the plasma at some value of s > 1 (but not too much).

There are users interested in being able to provide an alternative definition of the first wall, including an (r,z,phi) cloud of points.

One known issue is to consider how users will want to logically define component thicknesses relative to such a different first wall definition.