Add bg_intensity_sys_err to Lo L2 maps

[vineetbansal]

Closes #3559

We derive the background systematic error from the geometric factor's measurement uncertainty (the new GF_Trpl_H_unc column in the v006 ancillaries), as I_bg * sigma_G / G, matching Eq. 7 of the algorithm
document. Fill where the uncertainty is not given (ESA mode 1).

[vineetbansal]

Both the hydrogen and oxygen GF ancillary files have been bumped up to v006 (from v005 and v004 respectively).
The hydrogen v006 only adds a new column wrt the existing v005 version, as per the diff screenshot below:

The oxygen v006 is identical to the existing v004 version

While the introduction of new ancillary CSVs in the code (especially since they are identical to the old versions, as in the oxygen case) is less than ideal, this problem will go away once we move to true ancillary files (once issue #3536 is addressed).

[vineetbansal]

Update from Menlo from maps generated using code changes in this PR:

Thanks for putting this together. I believe these have everything we need for L3 at this point. We successfully generated spectral index maps from them.

[tmplummer]

Everything looks good to me!