As to the sound reproduction, I think, so far, Class D does not outperform class A/B and class A.
@niodari I've not found that the case at all. Class D has proved itself able to keep up with or surpass the best of class A or AB. Its all in implementation.
Over the past few years I've worked through a fairly wide range of amplifiers — Atoll, Akitika, Van Alstine, Sun Valley, and several others — each with genuine strengths and each ultimately pointing me toward a clearer sense of what I actually value in reproduced sound. When I landed on the AGD Audion monoblocks, something clicked that hadn't quite clicked before.
What the AGD does particularly well is resolve the leading edges of transients without hardening them, maintain tonal density in the midrange, and deliver bass that is both tight and texturally informative rather than merely present. Imaging is precise without feeling artificially etched, and the overall presentation has a coherence — a sense that the music is arriving as a unified event rather than as layered components — that I found absent or only partially realized in earlier amplifiers. I have a strong sense that the amps by @atmasphere would also fit this description, but I still need to try them.
But I think the more honest and useful observation is this: these comparisons only mean something once you've identified what you're actually listening *for*. Dynamics, precision, tonal color, soundstage depth and width, micro-detail retrieval, low-frequency control — these are separable dimensions, and different amplifier topologies genuinely do different things better. A well-implemented tube amplifier may do things with harmonic texture that solid-state doesn't. A high-bias Class A design may have a particular ease at low listening levels. The sweeping judgments — "Class D sounds sterile," "tubes are always warmer," "solid-state is always more accurate" — collapse distinctions that matter.
What I found is that once I was clear about looking for a certain (and different) cluster of priorities, the AGD wasn't merely competitive with amplifiers costing considerably more. In the areas I weight most heavily, it was superior. Other Class D amps I tried were not. What results is the *lack* of an overall verdict on Class D as a category. Instead, what I got is what atmasphere points out -- that one arrives ats a verdict on *some particular* implementation, evaluated against *my* preferences, in *my* system. Which is, I'd argue, the only verdict that actually means anything.

