fix(electrostatics): apply slab correction in Ewald and PME

[lil-lon]

Summary

• EwaldModel and PMEModel now pass pbc and slab_correction to ewald_summation and particle_mesh_ewald, so states with one non-periodic axis get the slab correction added in nvalchemi-toolkit-ops 0.4.0.
• Before this change pbc only reached the neighbor list. The reciprocal-space sum treated a slab as fully periodic, so energy, forces and stress included the interaction between slab images along the non-periodic axis.
• Added a test that compares both models with a direct Coulomb sum over the 2D lattice. This reference may be more thorough than the test suite needs, so I am happy to simplify it if you prefer.

Results change for slab states such as pbc=[True, True, False]. Fully periodic states are unaffected.

Found while investigating #606.

Checklist

Before a pull request can be merged, the following items must be checked:

• Doc strings have been added in the Google docstring format.
• Run ruff on your code.
• Tests have been added for any new functionality or bug fixes.

[CompRhys]

thanks for catching! in general my hope with #606 is that we can drop replace these electrostatics model's with MLIPops to become backend agnostic.