Interesting...but very incomplete.
He did add the caveat A "typical" pair of tower speakers....which then makes what follows somewhat true.
But he also states the conflict about localization (>80hz). Multiple low frequency sources (subs) can spatially average out the frequency domain issues to an extent (with the addition of EQ, make things smooth at LP and over a broader area). What about problems at 90hz? Or 200hz? Or 400hz? Can't spatially average out those, EQ "fixes" only for that point in space (which for stereo, may be enough).
Most (all?) of these type articles tend to neglect something else. Perception. What can/do we actually hear? What matters? Is "fixing" amplitude alone enough, when not using steady state tones, but instead, with dynamic signal envelopes...like music?
It is true that we can only "localize" to around 80hz, but the ear can detect phase changes (in/out of) down to around 40hz.
It is not difficult to find recorded music (especially with digital, acoustic genres) with stereo content <40hz. What will having multiple 80hz sources intended to smooth only amplitude do there? For a more in depth read of these issues:
http://www.davidgriesinger.com/asa05.pdf
For 2ch classical/jazz etc enthusiasts, things may not be so simple...if achieving high bass SQ is desired. IMHO.
cheers,
AJ