Improved scattered light model

[rcooke-ast]

At the moment a very crude model of the scattered light is implemented, and this is only used for KCWI reductions. Ideally, we should use the inter-order regions (for Echelle reductions) and the interslice region for SlicerIFU reductions. This enhancement would require that we pass the Slits calibration files into rawimage.py (because the scattered light should be removed before the flatfielding frames), and this is not currently well setup for this.

The current simplistic model for scattered light subtraction (at least for KCWI) was implemented in PR #1661. I assign myself to look into a better model for KCWI. Part of this effort should involve using the Slits information.

[kbwestfall]

Partially addressed by #1704 . @rcooke-ast , can you comment on what's left to do?

[rcooke-ast]

Yes, this is now "fully" addressed. KCRM is not implemented, but KCWI and ESI are both implemented. So, this can be closed.