Ground-up overhaul needed to conform all fundemental matrices to a single, coordinate invariant convention #233

PR #225 starts this by converting the coordinate dependent energy eigenvalues/modes of DCON to coordinate-independent eigenvectors with eigenvalues of energy per power-normalized-flux.

All fundamental matrices should be consistently formed in the same space. This includes, for example, the Permeability and reluctance matrices of PerturbedEquilibria. Of course, the resonant coupling matrices should be as well. Some thought should be put into the left hand side normalizations of resonant coupling matrices (flux is explicitly ok by [Park 2008], but check the others - shielding current, delta', penetrated field/flux, etc. - and add annotations confirming they are all also ok).

When implemented, the torque response matrix should be similarly normalized.

Lets make the code operate, as much as possible, directly in this consistently defined space.

The exception, for now, is the ForceFreeStates Euler-Lagrange integration done in displacement-space. After everything else in in a power-norm-flux space, we can close this issue and open a new you if we want/plan to attempt converting that to operate directly in the power-norm-flux space. The expectation is that the (m-nq) factors converting from displacement to flux will improve large solution numerics near rationals but possibly result in lossy suppression of the small solutions.

This should go on the back-burner lists for @ebursch and @matt-pharr - good weekend task for vibe coding