Is the conventional interpretation of the anisotropic effects of C=C double bonds and aromatic rings in NMR spectra in terms of the π-electron shielding/deshielding contributions correct?

Based on the nucleus-independent chemical shift (NICS) concept, isotropic magnetic shielding values have been computed along the three Cartesian axes for ethene, cyclobutadiene, benzene, naphthalene, and benzocyclobutadiene, starting from the molecular/ring center up to 10 Å away. These through-space NMR spectroscopic shielding (TSNMRS) values, which reflect the anisotropic effects, have been broken down into contributions from localized- and canonical molecular orbitals (LMOs and CMOs); these contributions revealed that the proton NMR spectroscopic chemical shifts of nuclei that are spatially close to the C=C double bond or the aromatic ring should not be explained in terms of the conventionally accepted π-electron shielding/deshielding effects. In fact, these effects followed the predictions only for the antiaromatic cyclobutadiene ring.