It’s hardly a matter of belief: we’re pretty sure of what the limits of human hearing are (and that can easily be tested) so if two devices are performing well above what we’d be able to discern in their performance you won’t notice a difference.
If this was not the case it would be in the best interest of any of the manufacturers of expensive equipment to setup a credible A/B test that would show how a difference can actually be discerned. But they won’t, because it can’t.
If you got a dud send them back, no point trying to “fix” it.