BUG - Bad angle formatting in GLOWS L2

[maxinelasp]

Parent issue: #1276

The spin_angle variable in DailyLightcurve (glows_l2_data.py:86) is computed by averaging imap_spin_angle_bin_cntr across all L1B timestamps. This produces IMAP spin angles, but L2 should use the GLOWS position angle (ψ_PA), which is obtained by converting from the IMAP spin angle using the position angle offset (Eq. 29/30 in the algorithm doc, Section 10.6).

Fix spin_angle in DailyLightcurve to store GLOWS position angles rather than raw IMAP spin angles. The conversion subtracts the GLOWS effective angle and applies modulo 360.