I think Bounder got it exactly right:
I believe that, as practiced by some posters on SF, dressing is a bona fide art form. I think, though I cannot, of course, prove it, that there is a germ of objective correctness in the rules of classical dress in the sense that many of these rules evolved for good reasons that speak to something fundamental about humans. But that is neither here nor there. All that is necessary is that there be a consensus that something looks good. We can then have an interesting discussion about why it looks good and how to replicate those essential elements, whatever they are, in other contexts.
This is what Manton and Vox are doing. Manton wrote an entire book about why the rules of dressing are the way they are and how to use them effectively. The premise for this book was that classical dressing looks good, for whatever reason. If you do not accept this basic premise, there is no point to the rules. But if you do, the argument Manton makes is, "People think this looks good. The reason it looks good is because of X. Therefore, if you do X it will also look good in this other context." He is not relying on "moral credibility" but on results.
So the "credibility" that Vox and Manton have comes from the results that are delivered from the theoretical structures that they propose. This is the "coherent set of principles" that Fuuma suggests. Note that this is an intellectually difficult exercise and goes very far beyond "Nice fit!" or even, "I think it would look better with a burgundy tie." Creating these rules does not rely on the eye or even the opinion of the "expert." Rather, it relies on the expert's knowledge of a great number of individual cases and his ability to formulate general rules from them.
To put it another way, Manton and Vox engage in a process of inductive reasoning that allows them to formulate a set of rules/principles. Once they have done so, non-experts may use these rules to reason deductively to produce a particular sartorial effect. Whether these rules are accepted does not depend on how many bespoke suits they own or on how many posts they have but on how well their rules work to produce the desired effect, whatever it is.