We should double check the units of the BT normalized overlaps

[logan-nc]

The sqrt(A) weighted space of the resonant coupling matrix creates an overlap with units of T•m. This is discussed in #185.
Normalizing this by Bt leaves units of meters, instead of being unitless. We should,

• Make sure any such normalizations done in the code are done to be unitless
• Check if all sqrt(A) normalizations use the A of the control surface, as per the [Park, NF 2011] physical description of an area integral over the control surface integral. If the areas of the resonant fluxes (singflx) are local to the resonant surfaces, then the dot product with the external field decomposed with the control surface area weighting becomes physically hard to interpret.

[parkjk]

Indeed an interesting and potentially important check. Can you check if Phi_x is already normalized (by sqrt(area))? I wonder if Park's corrigendum is still not fully addressed;; while we have been already using normalized quantity.