You should have seen me refer to sound as energy many times on AH and that the interference from collisions between modes can be constructive or destructive. It can also be more neutral because the interference will have whatever effect it can, depending on where in the phase the energy coincides but it's still considered 'constructive' or destructive'. Think of this as if the sound was sine waves- the modes would be complete waves at the frequencies that are determined by the dimensions (c=wavelength/distance) and if the wave from two modes intersect at their maximum, it will result in a positive peak, if they intersect with one at max and the other at min, they cancel. If one is at max or min and it intersects with one at 0 (midpoint), its effect will depend on whether the amplitude is min or max. Then, the 3rd dimension needs to be considered. In two dimensions, this can be seen in water waves when two points of surface interruption exist- some wave become larger, some smaller and at some points, the surface will be the same as it is outside of the disturbance.
If you know boaters, you might hear them talking about standing waves- this is where that term came from- it's a literal name for the waves that are abnormally high. These are caused by the waves coming in and hitting shore, then reflecting outward and if the results from the intersections are bad enough, boats can sink.
These examples describe the reason EQ can't completely 'fix' room mode problems. If it could shift the phase of the signal, it could, but only if it can do this acoustically and it's not possible to do that for only one room dimension. It can be done electrically- that's the reason for the rotary phase control on some subs and the distance control in AVR/AVP menus.