Co-addition of SCI1, SCI2, SCI3

[awhoward]

Let's develop a method to add SCI1, SCI2, and SCI3 together in a smart way. Smart here accounts for: 1) the fact that the slices have slightly different wavelength solutions, 2) cosmic rays in pixels of a single order can be identified and removed during co-addition (with appropriate modifications to the variance arrays), other oddities with our spectra.

This is partially motivated by the EPRV-RCN effort, but we desire this data product regardless.