Simplify logic for glc_renormalize_smb

[billsacks]

Description of changes

Remove the glc_renormalize_smb option on_if_glc_coupled_fluxes. This had been the default. Change the default to 'on'. This maintains consistent behavior between different configurations.

Specific notes

Contributors other than yourself, if any:

CMEPS Issues Fixed (include github issue #): Resolves ESCOMP/CMEPS#133

Are changes expected to change answers? YES: changes answers by greater than roundoff for T compsets (active GLC with DLND) (now SMB is remapped with renormalization for T compsets by default); answers should be bit-for-bit in all other configurations (unless there is some configuration/combination that I'm not thinking of...)

Any User Interface Changes (namelist or namelist defaults changes)? Removes on_if_glc_coupled_fluxes option to glc_renormalize_smb

Testing performed

Please describe the tests along with the target model and machine(s) If possible, please also added hashes that were used in the testing

Tested in the context of cesm3_0_beta02: ran:

• SMS_Ld5.f10_f10_ais8gris4_mg37.I1850Clm50SpGag.derecho_intel.cism-test_coupling
• SMS_Ld5.f10_f10_mg37.1850_DATM%GSWP3v1_CLM50%SP_SICE_SOCN_MOSART_DGLC%NOEVOLVE_SWAV.derecho_intel [Like an I compset but with DGLC%NOEVOLVE]
• SMS_Ly2.f09_g17_gris20.T1850Gg.derecho_intel

All tests passed; the T compset test changed answers as expected but the other two were bit-for-bit.

[billsacks]

@mvertens @whlipscomb @Katetc - this (finally) resolves #133 in the way we discussed there a few years ago. Note that this will change answers for T compsets, because now the SMB renormalization is always on by default (but it can still be turned off if desired).