MAINT: Hard failure for non-centered molecular force fields

[HaoZeke]

Basically, NWChem SCF (and ORCA probably) do not accept periodic boundary conditions. This is a bit a of a problem because incorrect inputs (e.g. non-centered in box molecules) can basically be ripped to shreds by the PBC respecting eOn optimizer steps. The fix is current on the user, to make sure the inputs are OK, but really should be done in eOn.