Here is a realy good 3-part article on DAC implementation.
Measurements can tell us pretty much everything we need to know:
Linearity - is the frequency response flat to within a fraction of a dB across the audible range?
IMD - does the filtering arrangement cause any intermodulation artifacts within the audible band?
Noise - what noise is there on the analogue output
Distortion - what does the distortion look like in terms of level and is it 2nd order, 3rd order etc.?
Multi-tone test - are any spurious inter-tones generated between the fundamental tones? Product of IMD.
0dBFS output level - should be ~4 V balanced, ~2V single-ended. Is the signal level high enough to drive the full output of your amp? Not too high that it saturates the input of your preamp/poweramp. Some DACs run ‘hot’, i.e. higher than these values. Back to back with another DAC, they’ll sound subjectively better because the playback is louder! This is why output level matching by measurement is vital when comparing different DACs
SINAD - how far above the sum of the noise and distortion is the signal? It’s generally accepted that better than 110dB is audibly transparent.
Output impedance - how low/high is it? Is it frequency independent? This will tell us if there are likely to be any effects on frequency response downstream.
Square wave response - looks at the effects of the DAC filters in the time/frequency domains.
This is sound, science/engineering based criteria on which to measure a DAC’s perfomance. No flat-earth doctrines here.
If two DACs measure closely enough that the differences are below the audibility threshold, then to all intents and purposes, they will sound the same.
Arguing about this stuff is like arguing the case for creation, when science demonstrably proves the case for evolution.