Why we blindly believe that if something measures good sounds also good and nothing can be improved in terms of sound quality? It would mean that the reason why Levinson 30.5 sounds better is only associated with added distortions. How about separate power supply with separate toroid's per section, how about class A output stage, better chassis / etc.
For over 25 years, I have been running a project with a few developers, engineers, and measurement technicians in which we always compare equipment first through very rigorous, strictly level-matched, and completely blind listening tests, and then investigate the audible differences using measurements.
This has involved hundreds of amplifiers and DACs across all price ranges, extending up to the mid-five-figure bracket. These included various expensive Class A power amps and many truly high-priced power amps and DACs from renowned audiophile manufacturers.
I have since parted with a large portion of my expensive gear, keeping only a few special collector's items.
Personally, I couldn't care less whether a device in my setup costs €50,000 or €500; for me, the sound is what matters.
In my main setup, depending on my mood, I run either an SMSL D200 or the original Cambridge Audio DACMAGIC 1 paired with a Yamaha A-S2100 or a NAD 2200.
However, I listen via the fully digital Sabaj A30a much more frequently.
My desktop setup includes, among other things, an SMSL D1 and a 3E Audio A5.
You might want to think about why someone who cares exclusively about sound and precise reproduction doesn't use devices that cost 100 times more to listen to music.