Add difference between atom and lambert to atom_scanner

[ennehekma]

Assuming Atom gives a more realistic a comparison can be made with the current deltaVs computed with Lambert . If the difference is calculated and stored in the atom_scanner mode plots can be generated, similar to spg4_scanner, indicating the error in deltaV that is happing from using Lambert instead of the more realistic atom.