Dare I throw a few simplistic assumptions up to be shot down in flames?
If we could cancel only reflective waves at the listening position we would have the equivalent of an anechoic room at low frequencies instead of a highly reflective one.
It would probably sound like being in a closely packed conifer plantation but even less interesting.
Fortunately we have a simpler situation in our humble HT/Stereo sytems.
It is assumed that the listener remains in one place. This avoids much unwanted complexity.
It is assumed that if we use equalisation it will be used to control only noticeable peaks and only at the fixed listening position.
REW will offer greater sensitivity to direct acoustic signals due to the very small delay (time=phase) between transducer excitation and detection.
Longer sinewave test tones read from our humble RS meters will average direct
and reflected waves at the listening position.
We can arrange the subwoofer to be at a roughly equal distance from the speakers.
Given the very long wavelengths involved the phase errors of the direct wavefronts are diminished to almost zero.
Phase errors exist for all frequencies in all positions in a room except for direct wavefronts reaching the ear. The human ears can cope with out of phase audio waves from multiple reflections. Because we've beeen doing this since birth (in caves of various forms of sophistication) for several millions of years.
I am really not sure that the average human listener would prefer a perfect, fullrange, point-source transducer in a perfectly absorbing room. (or its electronically filtered, pasteurised and equalised equivalent)
If semi-anechoic conditions were so desirable we'd all be outside watching films and listening to music in our yards and gardens. Fortunately (for all of us) the great indoors offers rather more than mere shelter from the elements.