Fix unit handling in ERF DNB normalization's saturation correction

[djhoese]

@kathys noticed that when using the saturation correction option on the dynamic_dnb (ERFDNB) composite that the result was all black. This was a dumb logic bug by me when I ported the code to satpy from Polar2Grid and the tests not being very comprehensive. This PR fixes it. See related PR and issue discussion in Polar2Grid: ssec/polar2grid#424

• Tests added
• Fully documented

[codecov]

Codecov Report

Merging #2044 (1e1d549) into main (4014846) will increase coverage by 0.03%.
The diff coverage is 100.00%.

@@            Coverage Diff             @@
##             main    #2044      +/-   ##
==========================================
+ Coverage   93.66%   93.69%   +0.03%     
==========================================
  Files         282      282              
  Lines       41995    41990       -5     
==========================================
+ Hits        39336    39344       +8     
+ Misses       2659     2646      -13     

Flag	Coverage Δ
behaviourtests	4.73% <0.00%> (+<0.01%)	⬆️
unittests	94.24% <100.00%> (+0.03%)	⬆️

Flags with carried forward coverage won't be shown. Click here to find out more.

Impacted Files	Coverage Δ
satpy/composites/viirs.py	89.86% <100.00%> (+3.46%)	⬆️
satpy/tests/compositor_tests/test_viirs.py	100.00% <100.00%> (ø)

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update 4014846...1e1d549. Read the comment docs.

[mraspaud]

LGTM!

[coveralls]

Coverage increased (+0.03%) to 94.179% when pulling 1e1d549 on djhoese:bugfix-erf-dnb-satcorr into 4014846 on pytroll:main.