There are a large number of videos on the internet where a person will claim that a component does or does not matter, that two different things are actually equivalent, or the opposite that two things that claim to be the same are indeed very different, or that they have "debunked" some long-standing "myth" about gear or audio or recording, etc. Often these videos seem somewhat scientific, or present recorded samples that seem convincing.
Most people will be familiar with this kind of "testing," we all do it. Countless hours with friends trying to figure out what does and does not matter in your tone. It's a quest, we're all doing it, and to a certain extent you get to make your own rules. I only put "testing" in quotes because the scenario almost always leaves something to be desired from a scientific perspective. Whether it's feeding an amp with a heavily truncated signal from a common digital looper, or making judgements after a mic/preamp combo that isn't flat-response, it's just not good science.
What I think is important to state right up-front is that if anyone doing this kind of low-res testing finds something useful for themselves, I think it's great. No one has the time to ferret out every aspect of electronic design, or audio design in general, so we all have to choose which rabbit-holes we're going to go down and how far. There was a point in my life when I was worried about guitar picks, for instance, after being at a party with two amazing guitar players trying out different esoteric picks. I spent a few months buying everything, carefully listening, trying out on acoustic vs. electric, clean vs. dirty, etc. etc. I reached some conclusions and moved on, I know my personal favorites now and when I think it matters most. I didn't get a durometer, or a spectrum analyzer with a measurement mic, or build a guitar-playing robot arm. But what I also didn't do is tell everyone that I had solved the myth of guitar pick sound. This matters a lot to me personally - Claiming that you have solved an age-old problem of an extremely subjective matter and broadcasting this as far and wide as possible, without actually having the findings documented, peer-reviewed, etc. It's just clickbait at that point and I think it does more harm than good to another individual on their own tone quest.
For instance, absolutely power amp tube type matters. It determines wattage, power consumption, impedance, it interacts with the power rails, etc. etc. It will determine some aspects of the amp that might not be easy to understand for a layperson, but they matter. Each tube type does have a certain sound, even if it is subtle, and variations within the type can also have a serious impact. Now, a single person trying to prove that it doesn't matter may be able to set up an A/B where the difference is not easily audible. That actually means nothing to me - it's like when you see A/Bs of pedal clones vs. the original, and the tester finds the settings that match best and exclaims "AHA! Gotcha!" Rarely do they sweep the travel of the knobs, evaluate the outside settings, really try to focus on where they are different. Doing that, I think, would be genuinely useful. There is more than one way to implement a design, so the fact that one design may minimize the effect of a component swap, another may take advantage of a specific aspect to the extent that any substitution will sound wildly different.
I am not a person who believes there are magical components either - A capacitor does not sound good because it is painted orange, and there is no reason to think a new capacitor sounds like an old one because they are both painted orange. To the extent that you can figure out why certain ones sound better, that is useful info, but again we all decide for ourselves how far to go and none of us should claim to have solved the myth of capacitors without getting a gas spectrometer involved. If you like the orange ones and decide to accept that and move on, I absolutely respect that. But let's stop short of claiming that "the orange ones are better."
What's really challenging to me about these videos is the insistence that it's a yes/no question. Does the capacitor matter or not? Well, neither and both are true. Designs, circuits, devices, these are cumulative systems. A single low-noise resistor may be hard to show the importance of, because the effect is cumulative. Using capacitors with better phase response or ESR is also cumulative. Swapping a single component in a system and not hearing a change does not mean the component doesn't affect the sound in some way, or have a purpose. Especially if that judgement comes after a few more stages, or in a recording. Equally, changing a single component and hearing a wild tonal shift would not prove that it's only that component making the change - it could be affecting a filter, bias voltage, any number of things. Would it be so hard for either side of this argument to accept that everything matters, if only a little bit sometimes, but nothing is magic?
One of the stories I was told that I really love, although I must admit to not knowing 100% if it is true, is about electrolytic capacitors. Over time, every electrolytic capacitor will wear out. It will need to be replaced. But there are those who say old caps sound better. And in some cases it would seem correct - when I did bench repairs, original stuff, from the 60s and early 70s, when working properly, seemed to sound "better" than recapped stuff. An older and wiser tech, I think it was Paul Wolff, told me that in the 70s the type of goo inside electrolytic caps, basically industrial Gatorade, changed. The Gatorade had been determined to be highly toxic by a new standard, so the chemical companies changed the formula. By all the normal measurements important to chemical companies, it was almost exactly the same. Capacitor companies started using the new goo, and updated the data sheets. Almost everything stayed exactly the same.
Except - The ESR was higher at low frequencies. Now, you may not know what ESR is - Equivalent Series Resistance. It basically describes how a capacitor acts like a resistor, even though it isn't supposed to. It is a small amount, but it also is frequency and voltage dependent. The resistance will change with voltage and frequency. So, without going into a ton of electronic design, in some cases putting a new cap in an old circuit, using the same stated value, the sound would change. Low end would suffer mostly. A person would put a new cap in an old circuit, and it would sound worse. "The new ones are worse!" Another tech would put a new cap in a different old circuit, and it would sound better. "The new ones are better!" Both people are actually incorrect. Once you know the root of the problem, you can just inspect the circuit, properly rate the new replacement, and everything goes back to normal. It's just a question of understanding what has changed and why, under what circumstances that will make a big enough difference that you care, then you can solve it without resorting to superstition in either direction.
So, old caps are not better, but old electros have different properties than new ones and the capacitor company may not assume it's important info because they sell to designers and engineers far more than repair techs and tone-seeking musicians. They assume you are going to read the data sheet and determine what value you need, rather than just grabbing a replacement based on the labeling.
Another similarly amazing story is about the transformers in old Neves - They were originally made by one company, which went out of business. Two companies stepped in - one bought all the machines and the other hired all the workers... both claiming to make the exact same transformer as the original. The good news is that I can never remember which is which, because all three make fantastic transformers.* If you accept that everything matters in electronic design, but that some things may only matter a small amount or under certain circumstances, you can short-circuit a lot of the nonsense. The most important thing, to me, is to follow your own journey, learn as much as you are willing to, and don't fall for anyone who tells you they have solved it quickly.
There are countless things like this in audio - AC power cables are often the source of myths and opinions, some people believe they don't matter and some people spend tons of money on fancy ones. What to do?! Well, it follows the same course as above. Let's ask instead, "Under what circumstances could it matter?" A tube amp, guitar or hifi, pulls a fair amount of current from the wall. As you pull more and more voltage and current over a wire, the gauge matters more and more. Because many tube guitar amps are old, simple circuits, they are very sensitive to noise on the AC power lines, so shielding on the power cable could matter. This doesn't mean that it always matters to you in your circumstance, but it means that a big hefty tube amp may show some actual sonic improvement when you use a heavy, shielded AC power cable. What this also tells us is that the AC power cable feeding your switch-mode based pedal power brick is very unlikely to matter. No magic, no claim it never matters, just an understanding of when it might matter and what you could do about it.
I try to design and build high-quality devices based on the things that are most important to me. It's the culmination of everything I've learned so far as a musician, guitar tech, audio engineer, studio and bench tech, and now equipment designer. I'm not here to tell you anything is magic, or that anything isn't. While traveling your own tone-journey, please reference my throw-pillow:
"Thinking components don't matter is as foolish as thinking components are magic."
*Correction 10/4/22 - KMR Audio published a great interview with AMS Neve's Joe Heaton which appears to debunk parts of my transformer story, and is also super interesting. You can also visit the AMS Neve page about this.