I'm going to attempt to boil one side of the argument (as I read it) down to its essence and where that seems to leave us:
Measurements tell us that most electronics beyond a pretty basic level of design are essentially competent and blameless, bar a few outliers. So there is not much to choose between electronic sources and amplifiers, so spending above a notional threshold where this competence can be taken as a given, is not justified by the performance.
Measurements tell us that, once above a nominal cost threshold, all cables measure similarly and will be sufficient, so spending any budget beyond that necessary for a basically functional cable is unwarranted by the performance.
This seems to rule out any spending beyond, effectively, budget equipment and eliminates any justification for upper or high end equipment on performance grounds. Spending beyond these modest requirements is therefore only justifiable on other grounds, such as functionality, aesthetics, build quality, reliability, etc.
It's a very utilitarian approach, though not really a Utilitarian one.
Utilitarianism does seem to take into account other aspects such as pleasure and happiness, and good to others (such as for example, creating employment for providers and makers of goods).
The corollary to all this is that any perceived benefit beyond what the measurements imply is argued to be down to, as Keith suggests, 'cognitive biases'. (Not necessarily 'expectation bias', which is a misapplied term for a quite specific form of experimental bias but it's crept into the objectivist lexicon so we're stuck with it, and its meaning in this context is generally understood so hey-ho).
