You are attributing the difference to something that may in fact have nothing to do with it.
I have proven this several times during design prototypes where I have taken IDENTICAL preamps and mated them up to different power amps/power supplies and when matched up correctly resulted in indistinguishable bass amps under double-blind testing. Therefore, it's easy to confidently conclude that if no difference could be detected then there was no meaningful difference present.
Now there are ways to INTENTIONALLY make a power amp behave differently, but that is an intentional design approach to achieve a specific effect. For example, I have done exactly this when developing the damping control circuit in the WD-800 (to emulate a specific characteristic or design quirk present in the walkabout's power amp) and also in the TT-800 where I added a circuit that tracks the constant damping factor characteristic of the Bass 400.
When I say that class D and class AB can be made to sound virtually identical, I am not talking about tube power amps, nor am I talking about power amps that have particular quirks or tricks built in to create an effect or a particular characteristic other than a simple power amp. I have been working with and designing with this technology for many years, I hope you can appreciate where I am coming from.