I think looking at Fig.8 can help. On the absorption spectra you clearly see that the maximal absorption peaks are at around the same intensity/height. But when looking at the lower spectra of the absorption derivative, the line that crosses zero in the isotropic case is a bit longer than that for the largest line in the axial case for example. So if you read only these latter spectra, you would think that the absorption for the isotropic case is more intense, but we can actually see in the absorption spectrum that this is not the case! Thus the line that crosses zero in the absorpion derivative spectrum can give a wrong impression.