Just on quick reading it seems like a load of nonsense; it ultimately comes down to "what sounds better" to some 'philes in general and this author in particular.
I am not a scientist or engineer so I cannot and will not offer a detailed technical critique of the article. However:
This guy and other 'philes seem to be bent on discovering and isolating an "x factor" beyond detection by standard (and supposedly insufficient) measurements that accounts for the allegedly superior performance of their pet topologies, and for the sonic refinements they claim to hear. Remember, these sonic differences tend to disappear in ABX (double-blind) tests of all but the more "eccentric" circuit topologies like single-ended triode tube amps. The "AES types" and engineers/mainstream scientists in general are often held in contempt by these folks. To me, it sounds a lot like the perpetual motion machine types who castigate "conventional" scientists for their "blindness".
If the presence of such an "x factor" in either electronics or human hearing were seen to be a real possibility, one would think that the much-derided mainstream "AES types" (not to mention audiologists, psychoacousticians, and maybe even physicists) would be falling over themselves trying to find it, develop measurement protocols for it, and design new eqipment to exploit it. After all, it would confer considerable prestige on its discoverer and very likely considerable profits on the manufacturers who patent and exploit the novel circuit topologies resulting thereby.
Remember: for all the huffing and puffing by this author and others, the standard for music reproduction is and always has been that the output signal should resemble the input signal as closely as possible. Current measurements are capable of detecting deviations from accuracy far below the thresholds of human hearing, and modern solid-state electronics deliver accuracy well within the limits of human hearing. Tube gear can, too. But the single-ended triodes the author is talking about generally do have audible distortion and frequency response that is far from flat. Not to mention wimpy power output, like 9 watts/channel! Distortion of any kind is just that: distortion, or a deviation from fidelity. It might even sound "nice". Still, as the notorious "AES type" Doug Self has pointed out, why don't makers simply add a "niceness knob" to add whatever kind or degree of harmonic and other distortion one desires?
Anyway, one particular type of harmonic distortion (I can never remember if its even- or odd-order) does sound rather nice ("euphonic") because it adds harmonics an octave above the fundamental, adding a bit of brightness and (to some) the appearance of added "detail" and "air" and the like. My Audio Wisdom page links to a site where you can hear samples of even and odd order harmonic distortion for yourself.
Bottom line: even though he sounds techy and does all sorts of measurements, it still sounds like pseudoscience to me.
[EDIT] I see the author has a techical background and worked for Tektronix (a respected mfr. of oscilloscopes and other test equipment). So you're going to say, "ha, Rip! The guy's a techy himself so who are you to call it pseudoscience, Mr. Smarty Pants theater major?" To which I reply: the annals of science (and other intellectual pursuits) are full of otherwise brilliant people who forget their intellectual self-discipline and become quite dotty when they stray outside of their specialty, even if it's "right next door" so to speak. It is one of the mysteries of human folly.
If he does have real technical or scientific insights, why not publish them in a peer-reviewed technical or scientific journal where they can be critiqued, tested, and confirmed, refined, or refuted by other scientists? That's how science works, folks.
I leave it to any actual scientists and engineers here to offer a detailed technical critique of the article.