Improving generality and speed of newly-ported MESA code

[nriemer]

• Check if Taylor-expansion coefficients can be replaced by yaml files
• Check if hard-coded Taylor-expansions multiplies can be replaced by matrix operations.
• Relax hardcoded species indices
• Figure out where the hardcoded liquid-liquid temperature-dependent thermodynamic parameters come from
• Reevaluate solid and liquid phases separately (aerosol and electrolyte; e.g., aero can be a view of 1D liquid first and then solid)
• compute_activities (maybe other functions) need aerosol state, phase, and hysteresis leg set beforehand, so make sure these are carried correctly
• Consider adding update_thermodynamic_constants instead of directly computing constants within compute_activities (making sure that temp. is handled)
• Fix duplicated code to handle Fortran go to statements
• Check to make sure that the activity of water is being updated with relative humidity (i.e., aH20 should not be local in bin_molality and fnlog_gamZ)
• Fix hardcoded tolerance for ASTEM
• Make sure aerosol_water_total is an input where necessary and not hardcoded (e.g., in team_invoke)