Actually, I was going to go down the psychology path on this one. Expectation bias, etc. This is why fancy cable makers abhor double blind listening tests because 99% of the time they reveal charlatanism.
I've used stuff like the "jitterbug" (ironic name because it can't possibly affect jitter) to prove my point many times. What I do is simply alternate then skip one switch, so the cables/devices are then the opposite. Almost every time, the listener claims the benefits of the product they THINK is playing! That's no joke. It's actually difficult to reveal the results because people don't like being told they were fooled.
I think there are a lot of genuine bogus products that don't cause any change. In those scenarios, sure, the manufacturers would not prefer anyone do any such testing and in fact they will avoid discussion of the tech, specs, and measurements. That's certainly one group.
Although, I think amongst the collective of sensible customers and manufacturers who've been doing this for a long time, they generally do put out non-bogus products (i.e. they cause a change) and the customers generally can tell if they are certain YES, this is doing something vs. "guys, I think or feel like it may be doing something...but I think I'm shooting in the dark"
Perfect example (yet again) is of Jitterbug. I have 2 of these. Audioquest for example recommends that their primary use is in the chain (computer's USB output -> Jitterbug -> DAC). However, they've also recommended that you can plug in a second one to an unused port for further SQ improvement. I've always felt and written that Jitterbug in the chain with all of my DACs have caused some change in sound. However, when plugging a second one to an empty unused port and then plugging and unplugging it with the music playing, I don't feel much confidence that any change happens. Of course, not everyone will want to admit that or go through such experiments. Still, I think that amongst the more experienced and skeptical listeners/tech curious crowd, we do get a good handle on whether things make a difference or do not.
I've somewhat moved on from "whether things make a difference or not" from the early years of listening when I was still developing my listening skills and gaining testing experience. For the past few years, the more interesting for me has been "what does the difference mean"? In playback settings fidelity, neutrality, and accuracy in reproduction are a high priority for me. So when a product comes along that's supposed to be inert in its influence in the chain causes a change, I'm more interested in finding out whether the sound is now more or less accurate and whether this change toward more or less accuracy corresponds proportionally or inversely with my enjoyment or interpretation of SQ.
Make no mistake that I'm still skeptical of manufacturers (from boutique to even some big ones - including the 'big' one we have on this forum itself). I've asked them MANY times on this forum and elsewhere: why their products are sensitive to such changes and not immune currently and also what they could do to make things more immune to changes elsewhere. Responses can vary from "it shouldn't be happening", "I don't know", or "Use whatever sounds good"...rarely do I ever get a satisfactory explanation. BTW this wouldn't be a case of just myself imagining this. So many other customers would say the same thing and in fact the manufactures themselves will poll the customers what they think offers the best SQ or what's their "preferred" connection/arrangement (as to not make their product look bad).
It makes me question whether the manufacturers are aware of the full picture and are purposefully hiding the explanation to not make their products look bad, or whether they too are not fully aware of some of the aspects.
This is why I haven't bought a new audio product in over 5-6 years. So many new areas, such as network switches and stuff. Some will say that they make no difference. Meanwhile, other listeners and the manufacturer will say that YES there a difference AND it makes the SQ better. However, the problem arises when they can't explain the tech or provide any measurements. it's either 3 things:
1) They are knowingly lying outright.
2) They don't want to reveal their secrets, but are aware of the full picture.
3) They don't really have a full grasp of things.
I was astonished when I was reading some of the comments by the manufacturers of the EtherRegen during its development. One of them (superdad?) said that we don't fully understand all of it, but to our ears its an improvement over previous product, so we're still going to release it...I couldn't believe that. You are the manufacturer! You are the one group that should know. Otherwise, how can you say with confidence if whether any changes you cause are for the better and not for worse!?
I'll follow up on the blind testing and best practices for them, got to run for now.