I don't think there's anything new in this article. They're actually talking about the difference between Class A and Class AB amplifiers, most old valve amps being of Class A design and the trend with solid state amps being towards Class AB. Depressing that this made phys.org.
Very few classic tube guitar amplifiers were Class A, the vast majority being AB. Really the only ones that were class A are single tube amplifiers like the Fender Champ. The Vox AC/30 that everyone thinks is class A is just cathode-biased class AB. The Marshalls that Hendrix played through, the Fenders that Steve Ray Vaughn used, and the Boogies that Santana is known for, all class AB.
Sure, class A does tend to have more even-order harmonics, compared to AB, but even class AB tube amplifiers have more than any solid state one.
So... did these tube AB designs have particularly asymmetrical lobes to their gain curves? Otherwise I don't see where these even order harmonics are getting introduced.
Yes a bit due to asymmetry in the phase inverter circuit driving TD push-pull, but mostly because the odd harmonics are naturally cancelled out in the output transformer.
