Class B has clear discontinuity as one transistor gets biased with the other one not working. That gap is compensated in class AB by pre-biasing the two transistors so that they can hand off to each other quickly. That compensation is simple and hence the reason there is no such thing as class B audio in the market.
Class A's advantage is simplicity in circuit design. When transistors were expensive and so was labor, having an amplifier with half the parts was an advantage. That has not been the case for decades. The efficiency advantage of class AB has obsoleted class A in audio for years.
Thank you for the excellent explanation. I truly learned something here. =)
Though you heard no difference, is there a theoretical benefit to what Schiit claims to have done here, with extending Class A biasing up to 10W or whatever?