For each of us Ralph the truth awaits in different things. For me it’s in the listening to music and as I said to some mates recently the notion of connecting with musical truth. Semantic testing arguments aside I’m comfortable in the notion of each recognising their own rightness.
Everything I invest in this now is in the experience of getting deeper in the music. That is my goal. Fidelity isn’t a singular universal concept so we all apply it as it best fits our perceptions and values. I’m hoping you create a class d amp that could do that as well. But in the end it’s your call on what it should be as designer.
IMO/IME the stereo should not attract attention to itself- the music should simply be.
The trick is connecting the dots between what we hear as music, how the ear interprets sounds and finally what engineering is needed to match the ear/brain's perceptual rules. So it may well be that equipment that does that may not have the 'best' measurements. What the ear perceives as important can be different from what is considered 'good' or 'excellent' on paper, but only because the spec sheets tend to ignore human hearing perceptual rules in their design.
Of course we want the system to be silent, but it will have noise. If it has noise, it should be of the sort that conforms with the ear's exception to the masking principle- IOW the ear should be able to pick up detail below the noise floor. In those systems where the noise floor is not of this nature, they will be perceived as lacking low level detail for this very reason. This is the difference between a black background and one that is less than that, and for clarity the blacker background is the former rather than the latter.