* Add Chebyshev2D departure term type

* Split out reducing and mutants to their own headers

* Remove unused function

Add runtime test

Add test and expose in Python

Added a mockup of how one would hardcode models into the shared library (a la CoolProp), dynamic loads just need to be replaced with hardcoded strings, probably encoded as binary

Add tests for all binary pairs
Add conversion script for HFC mixture xi and zeta

Also allow paths to be optional

Previously you always got Lorentz-Berthelot no matter what

Not yet fixed for mixtures

Small computational speed bonus

Also not clear if this is the best way to include it.  Probably need the std::variant approach here too.