A person's knowledge of the inside workings of a machine is largely irrelevant because the ONLY thing that is being judged is the end result. A racecar driver doesn't need to know how the Internal Combustion Engine works and also be able to design said engine to be able to judge it on its performance on the track. Audio Equipment is designed to recreate the sound on the recorded disc. A great many audio companies hire fully qualified degree-holding engineers and use the best measuring equipment money can buy - even SET amp makers!
I can't speak for other writers. Most magazines hire a large staff of people and some are more objectivist and some are more subjectivist and that means that yes some will fall for what you would deem as snake oil like $1000 USB cables. In my case, I like SET amplifiers - which technically is inferior to solid-state - but I am shielded from the snake oil argument because a SET amp WILL be distinguishable in a double-blind test from a SS amplifier - therefore a difference DOES exist. The difference is established. So in a preference-based blind test - if one chooses the SET because it "sounds better" then it "sounds better." It is then a valid preference. The Audio Note J/SPe speaker was sent to Hi-Fi Choice and blind level matched sessions occurred with a panel of listeners - the AN J is a standmount and they put it in a group test of famous standmount speakers from B&W and others that probably measure better. The AN J won with ease and all the comments noted the bass response was prodigious. So because the Bass was so much deeper and better than everything else in the class - the magazine felt it was unfair - so they put it in a test against those competing manufacturers' floor standers with bigger drivers. The AN J beat all of them as well. The panel all chose the AN J - the magazine bought the speakers to be used as their reference.
Now had they just read the measurements without the blind test there is no way they would choose the AN J - the blind listening changed their minds.
Designers of anything should be doing blind level matched listening - when the famed John Curl designed a loudspeaker he noted that the thing measured fantastic - then he listened to them and disliked them and decided to go back to amplifier design because although he created a technically superb speaker it sounded rubbish.
Lastly, I am an Audio Note owner - they are one of, I should say the ONLY, company in all of audio, that designs and builds their own transformers in house, their own capacitors, resistors, knobs, chokes, circuit boards, casework, connectors, wiring. They are also making their own valves and recording equipment and ADC in the studio.
As a designer yourself you should look at the walkthrough of their plant to see the kind of equipment they have to play with. It's not just some snake oil guy in a garage - again measurements they aren't going to win the owner himself says that their DACs "measure like a bag of nails" but in their listening tests it simply sounded better. Then what? If A sounds better but measures worse what are you going to do? You have to make the thing that sounds better - that's the whole point of buying a stereo system - enjoying the sound not reading the graphs.
Part 1 is with Andy Whittle who heads Rogers Loudspeakers
Part 2 is with Audio Note's lead engineer Andy Grove - who worked with Quad and designed their Anniversary amplifier.