Loading CIF might not require symprec

[kbarros]

If the CIF file already has spacegroup info, then in principle we could use the spg_standardize_cell feature of spglib build a "standardized" crystal which will correct for small errors in the atom positions. If we do this, then we wouldn't need the user to provide a symprec parameter.

[kbarros]

This may have become easier now that we require the symops of a CIF to conform to an ITA standard, and now that we have Wyckoff position tables in Sunny.