Well, I have been listening to a lot of DACs and several amps. I find very large differences. Yes, the Benchmark amp is dead flat, super low distortion, but to me it sounds lifeless. Just for fun, I grabbed a Fosi V3 ( also super low numbers) and at very low levels, not bad. Bring it up to moderate levels and it was terrible, well within 10W range on my desk. ( My 2W Rekkr does very nicely). So relegated to my garage system. I find every class D amp either lifeless or something wrong I can't put my finger on, though they are getting much better. If return shipping was not so high, I would like to try a March. My JDS DAC sounds better than my Topping or Schiit with traditional numbers a good 50% "better." as well as several much higher priced SMSL/Topping/Cambridge I have sent back and the internal DACs of some almost prestige level integrated amps. I do want to play with a Qutest for a while. My specific area of concern is female vocal edginess. Not bass weight or some ethereal "air" I don't believe is in the recordings to start with. Maybe a Denafritz. Much worse traditional numbers.
Simple SNR/IM/THD numbers are valid, but only part of the answer. The missing component is how our brain processes the information. Also not determined by simple pure tone testing. Good data, but not the full story. With a speaker producing 0.2% third harmonic, explain why our brain likes or dislikes differences in electronics of .000x. It seems to, but we don't have an explanation. I propose masking is part of it. If numbers were all of it, no one would fool with vinyl. We do because it sounds better in the preferences of some people. Clear differences even in a tone arm and we have no idea what measurement describes that. Personally, I found surface noise and ticks and pops to be more objectionable than even second generation digital so I switched.
If you read my post more carefully, I am speculating on what DISTORTIONS possibly make our music more pleasurable and which make it less so. Tubes are a prime example. Distortion from moving coils another. High noise floor and limited dynamic range from Vinyl? (OK, most CD's are super compressed too thank you loudness wars) Something in most DSP I find objectionable. Testing through JRiver identified clipping due to filter overshoot to be one issue ( easily solved). Does pre-ringing mess up our perception of transients? Don't know, but an easily measured kind of distortion. Lots of issues to explore.
Again, if you have a better understanding of DACs and how digital filters work, you may not have the same " reads numbers" viewpoint. Reading numbers is only step one. Even the cheapest dongles do some DSP. The A2D has just as many issues as do all the processing done in the mastering. To avoid any DSP, well maybe an old Sheffield D-to-D lp. The King James Version being one of my favorites.
Where something is assembled is of course your choice. I prefer to select the product for it's merits. I am tempted to get a Vidar and see how it compares to my own MOSFET. So far few have equaled it. It benches pretty well too, better than I can measure through a Focusrite. ( distortion added by A2D). I am not wealthy enough to buy "prestige" amps. I am not convinced ( yes, I have borrowed some) are any better in any system I can afford. Drop me off a Levenson of Bryston if you wish :)
We can agree, many PAs are terrible. Many venues are worse, but I find the musical experience to usually be better. "Why?" is the question. We did walk out of a MoodyBlues concert once as it was totally unlistenable. Why does a Strat into an over-biased Fender sound great live but a recording of it does not? Why can't we seem to record a single classical guitar or piano convincingly? We don't know.
Yea, the Anthem DSP is decent for HT. Beats Audessy any day, though my old MXR310 is not exactly SOA amp wise. Have not had a MiniDSP with DIRAC to play with. ( digital in, digital out as their DACs are about an Apple dongle level). I do my eq on my main system by careful crossover design and let my brain sort it out without any processing artifacts. Our brain actually does a very good job of eq if it is not too far off. ( component burn in I propose is 90% our brain getting used to it, 10% actual technical)