fluxCal phase 2 vs phase 3

We make 3 steps in fluxCal focal plane fitting.

1. fit the normalization of the fluxCal vector
2. apply the focal plane fitting using an approximation to compute synthetic fluxes. This is fast but not exact.
3. do it again with the mathematically correct expression with the initial guess from step 2.

This ticket investigates how good/bad the 2nd step is. If it is sufficiently good, then we might skip phase 3, which is by far the most time-consuming part.