I think there is a big understanding between "Sound Quality" and "Sound Quality".
""Normal"" music listeners (Not Audio Scientists) mean "It sounds good" when they talk about high sound quality. Which is not the correct word, yes, but that is the easiest way for people to explain when they mean "Something sounds good". They say "Sound Quality" but often they mean the overall presentation of the sound rather than measurable sound quality.
Of course no human being will ever hear any difference between an device that does have an Dynamic of 105db and an Device that does have an Dynamic of 145db. The Human ear is just not good enough to perceive it so even though, when one is better on paper, nobody really cares. And that is exactly why nobody (except scientists) talks about that. Why would you?
What people care about mostly is how authentic the DAP can make things sound like. If you play an acoustic guitar and then record it, due to the limitations of recording (no matter how well it is done and how well it is mixed and mastered) it will never sound like in real life. Microphones are incapable in doing that.
The Headphones turn back the lifeless and thin monitoring sound into an realistic and authentic sound, and most headphones need a colored source to do so.
If you want that flat, straight out of mic/mixing/mastering Monitor sound, you only need a DAP when you need high output power. USB Dongles are more than enough most of the time.
But if you want an setup that turns this back into an real sound, an lifelike sound that sounds as close as possible to the original instrument (and not the the monitoring sound of the recording), you need sound coloration.
Sound Coloration is mostly done by big and expensive components like wound capacitors and large LCs. And they takeup a lot of space and need a lot of power cleanup to work well and even then, will degrade the measurable sound quality.
But of course they rather sacrifies 20db of dynamic to produce an better sounding music because, why not? Nobody will ever be able to hear the difference in dynamic so who cares.
And the DAP who can do that best is generally perceived as having the highest sound quality, even when its worse on paper than devices with much better measurements. And i this leads us to the initial issue.
Are all daps worthless shit because they measure bad?
They measure bad because their components are designed to colorate the sound in the best way possible, not to measure in the best way possible.
Plug the IER-M9 into an LG V60 and it will sound like crap. Plug the IER-M9 into an ZX507 (which measures way worse) and it will suddenly have an authentic, realistic, lifelike sound that tricks you into instruments that are really in your room.
What do you want? Its a matter of taste, a matter of decision. If you like boring and lifeless sound, the better measuring LG V60 is more than enough for you, of course. An USB Dongle is the best choice, but most people don't want that sound.
The best sounding device i have, by far, is the TA-ZH1ES. And it does have less output power than a lot of DAPs out there. Why does Sony build an 4.4kg stationary brick that can easily beat by an DAP in terms of output power and measurements? Because it sounds better. That is why.
The best portable device i have is the WM1Z. Its not as good as the TA-ZH1ES but it gets, considering its size, very close to the performance of the TA-ZH1ES. Its not there, but good enough for on the go. By using 13 wound capacitors and 6 large LCs (that would never ever fit in any Dongle) it tries to get as close to the TA-ZH1ES as possible. Its not possible because it is limited by space, but it gets very close.
If the WM1Z can't do it, because its limited by space, of _course_ and 4x2mm Sound Chip in a Smartphone can't do that. And of course an USB Dongle can't do that.
I never use my smartphone (with or without dongle) because every Headphone/Earphone i own sounds boring and lifeless with any Dongle i tested so far. But of course that is the case, Dongles are too small to fit components like wound capacitors or LCs and they are essential for "good" sound. Not good in terms of measurements, good in terms of how does the music sound that comes out of the headphone.
Even my girlfriend who does have damaged hearing due to medication and is not an audiophile by any means, instantly and without doubt hears the difference between a smartphone dongle, her old Walkman (ZX507) and her new Walkman (WM1A).
Quoting her, with the WM1A, for the first time of her life when she listened to her favourite song, it sounded like the singer is really here, in her room, singing in front of her. It no longer sounds like music out of an earphone and more like music that is really there.
That is the reason why you buy a high quality DAP, because you want that sound. If you don't care about that, you don't need a DAP.
That is why she thinks the WM1A is the best DAP even though, in terms of measurements, it performs much worse than her dongle in every single aspect (including output power).
I think its the false believe that people want to get as close as possible to the plastic and limited monitoring sound of the recording. No musician i know cares about that sound. That is the sound the people in the Mixing and Mastering room have to work with, that is not what you want to hear as a music listener (in general, there are of course exceptions, especially audiophiles and audio scientists).
I think limiting Audio Science to the best measurement on paper is just scratching on the surface. You want that your music sounds best, and then, when you archived that, you want to make it the highest quality possible.
That is why a lot of people dislike the THX AAA Amps. They measure exceptionally well, but they sound boring, lifeless, thin and as far from the real life sound as absolutely possible. Of course they do, they are achromatic amps, its their job to sound like that, but that is not the sound the people want.
Most people do know how instruments sound in real life, they do know how bands sound life when being on a concert, they want _that_ sound and that sound is only capable with sound coloration done by big wounded capacitors and LCs who do barely fit into DAPs and of course not into dongles.