I think that the computational results, in particular, the ionization energies(IE) and electron affinities(EA), are considerably improved.
However, I am just wondering why the authors did not recalculate optical properties with using TD-DFT with employing hybrid functionals.
If the authors understand why LSDA is not good for IE and EA, they must think that optical properties also should be computed with not LSDA but refined functionals: is there any reason why the authors persist in using LSDA?