Aligning how we define chemicalsystems in the end states

[IAlibay]

It would be great to try to enforce how end states are defined for different types of simulations.

For example, for the ABFEProtocol we currently define the end states as a stateA being a ChemicalSystem with all the Components and state B being the same ChemicalSystem without the disappearing ligand.

Here is an issue in gufe which explores this : OpenFreeEnergy/gufe#686