Remaining issues with validity metrics

[smglsn12]

TODOs for the validity metrics

The charge neutrality metric has been significantly updated with the push a few weeks ago to add an empirical calculation of plausible oxidation states (and determine whether a charge balanced composition is possible based on these oxidation states). This is draft PR #7. Finishing touches on this PR:

1. Add final validity calculation based on oxidation state score (set a threshold for now - probability of greater than 0.1% seems reasonable to be deemed a viable composition from a charge neutrality perspective)
2. Integrate this into benchmark workflow that has now been updated by the refactor of the stability metrics

Density metric
3. needs to be updated to have a moving window to check that there aren't any clumps of very high density atoms in the unit cell - this could be the case even if the overall density is quite reasonable

Lattice sanity checks - is a=b=c for cubic etc
4. I don't think this is implemented atm but will be very easy to do

Overall validity
5. Set up overall determination of validity - will have to clear all the metrics to be deemed viable

[smglsn12]

1, 2 and 5 have been done and merged into main