The acoustical effect of cancellation notches get overrated because they look so bad in the frequency response and CSD plots.
With normal music material you'd have a hard time to ABX a difference between one and the same speaker with such a notch (emulated with a proper filter) and without. The reason is such deep high-Q notches need tens of cycles of a sine with constant amplitude to develop their depth (and then release a "reverb tail" when the sine stops abruptly. You simply don't hear it with the normal decaying spectrum even when you hit the right note.
EDIT: In general, a seemingly ragged frequency response in all the fine-grain is completely unrelated to how it sounds. A known worst case are coaxials measured (and listened too) direct on axis, as the total symmetry emphasizes irregularities, a very inappropriate magnifiyng glass.
IHMO, these high resolution plots should be forbidden to be published as people will get them wrong and fussing about it.
Further, cancellation notches is one of the things we are so accustomed to, basically our whole sound source distance judgments in everyday life is based on it to a big part so we just don't bother (even though it is different than a resonance-based cancellation). If we did, the floor bounce of speakers that is ususally very strong would totally spoil our listening experience... which it obviously does not.