Apply sky subtraction, deduced from DCB to SuNNS data, for limited number of fibers

Apply algorithm introduced in PIPE2D-793 to real SuNNs data.

We want to test how the current implementation of Zernike-based 2d psf will handle SuNNs data.

The limitations will be:
1. only some fibers available
2. somewhat different pupil illumination of DCB data and SuNNs data
3. relatively poor fit in red, which in the meantime has been improved, but I have not generated full solutions

Nonetheless, this is critical for testing the implementation and identifying possible problems.