Implement settings thermodynamic equality

[IAlibay]

It would be great to have a way to check if two sets of settings are "equal", i.e. that they would lead to the same reduced potential.

One approach here would be to somehow "tag" settings on if strict equality is necessary, that way we can just loop over these when doing an equality check.

[dotsdl]

One way we could evolve the model in the direction of separating thermodynamic settings from sampling settings without breaking the gufe tokenization model is to make sampling settings something we supply to Protocol.create. This would make them equivalent to a kind of runtime setting, and wouldn't be baked into the Protocol instance's settings at all.

From an execution engine perspective, such as alchemiscale, this would mean that users would specify these sampling settings on Task creation. This does put some burden on them to do this as a separate step, but would gain this sampling flexibility across Tasks computing the same Transformation.