Further abstraction in ModelMethod

[JosePizarro3]

@ndaelman-hu @Bernadette-Mohr

After talking with @JFRudzinski, it seems we should be more consistent on treating contributions of physical properties, more importantly when talking about models (either FF or TB). There are cases where physical property contributions (e.g., the total energy of a system) should have refs to specific terms in the Hamiltonian or model solved.

For this, I think it makes sense to add a new layer of abstraction for ModelMethod, and add a repeated subsection called terms under it. This will allow us to add a method_ref to the base class of physical properties/contributions so that it points either to the total Hamiltonian or to the specific part of the Hamiltonian.