Do you have evidence of this, or is it an opinion?
Then how do you explain the 7.2 format?
Good questions.
I could not find much on the Net, so I just did an experiment.
My studio has a slight room dominant reinforcement right at 50 Hz, just 4 db, but enough to create a slight warmth. Luckily my speakers are critically damped so the speakers do not overly reinforce this dominant mode.
I have two audio oscillators. Normally below 80 Hz my system is mono. However it is easy at the flick of switch to change that. I set my prepro to pass through with no processing and all electronic crossovers after the Rotel pre/pro.
I set two asynchronous oscillators to output 50 Hz at exactly 1 Volt and connected them to the left and right channels respectively. In stereo the 50 Hz room mode 50 Hz peak was reinforced by 3 db compared to mono.
Now my mentors told me years ago that deep bass should be mono, and I plead guilty to taking it on faith until now.
What else do I know. Well it has been known since the stereo era that in live sound reinforcement stereo is hopeless as it reinforces room peaks and leads to feed back and therefore significantly reduced forward gain from the mics. So reinforcement is always mono. I used to do a lot of reinforcement of live gigs and can absolutely attest to that.
The next is recording drums. I have just reviewed that literature, and there is overwhelming evidence that drum rigs are much cleaner and reproduce better recorded mono and then panned. That was always my practice.
I see there is lot of debate on some other forums about stereo subs. The consensus seems it is worse. I think it might be coming about for satellite systems as they have to crossover higher. In that case stereo subs would have a definite advantage.
Bass should only be coupled below 80 Hz in my view, and I believe that is the correct approach.
At those frequencies bass does not localize, so I see no advantage to more than one audio channel in the last two octaves. Potential phase summing and canceling super imposed on room reflections seems to me to be ripe for trouble.
However other than my experiment tonight, I can not find an in depth analysis, so may be it is an opportune time to visit this issue.