Improve Elliptic Grid Solver

[juddmehr]

Currently, the elliptic grid solver defaults to applying a successive line over relaxation for several iterations, then applying a full newton solve to converge the grid down to 1e-14.

The jacobian for the grid elliptic system is very sparse, so we should implement a sparse jacobian for the newton method.

It may also be faster than a full newton to apply a quasi-newton method. This could also be explored, though the sparse jacobian should be done first probably.

[juddmehr]

This is perhaps unnecessary with the recent reduction in state size associated with fixing bugs related to ImplicitAD implementation. Though it may be still worth looking into a sparse jacobian down the line.