Interface issues in experiment drivers

[LenkaNovak]

Due to the changing interface and AMIP urgency, the interface of the main driver is not fully optimized. This issue keeps track of the TODOs.

• bring in interface features in src/ outlined in Coupler-specific Interface #44
• Store all global variables in appropriate places - date0, landmask
• add unit tests for the Time Interpolator Design #88 suite
• move AMIP/mpi_moist_earth/coupler_utils/ to src
• remove regrid_space (and use a re-ordered boundary_space instead), though this may require ClimaCore changes. (see Separate regridding from file reader #89 )
• replace swap_space! where possible - see comments in Interface changes pulled from #90 #93
• remove unwanted slab land model
• using macros vs functions for callbacks Macro vs functions for periodic coupler callbacks #95
• using mutable vectors for single floating points to make them mutable in static structs. Perhaps these could be replaced with functions? See Daily interpolation from monthly data (v2) #105

[LenkaNovak]

mostly complete, rest to be addressed in #205