There are some issues in there.
1.
No one here had said SINAD is about audibility. SINAD is always, always shown in the dashboard in my reviews where you see the full spectrum of noise and harmonics in the FFT. I routinely analyze that and comment on whether there are audible concerns. I have also talked about limitations of THD+N in my article I wrote for Widescreen Review:
https://www.audiosciencereview.com/...derstanding-digital-audio-measurements.10523/
"Below THD+N is the SINAD figure I referenced earlier. SINAD is actually the same value as THD+N but expressed in dB relative to the signal amplitude. This is good because we can, at a glance, compare the SINAD to our hearing dynamic range of 116 dB and know if we are better or worse than it. This is a back of an envelope computation for audibility so don’t completely run with it. Use it as a first order approximation. If a DAC has better SINAD than 116 dB, then we have very high confidence of transparency. If the SINAD is below 116 dB, then the nature of the distortion spectrum will tell us if the distortion is audible or not (i.e. we have to look at the FFT). "
As I have always said, when you achieve high enough SINAD, you get a free pass in having to learn psychoacoustics. If you do not, now you are in a hell of a situation having to analyze all aspects of distortion and prove such is inaudible. And for what? There is a ton of gear today that achieves my transparency metric. Why on earth do you want to dabble in more non-linear gear as to then defend it this way?
2.
The masking thresholds are for general population. While with linear distortions (e.g. frequency errors) we all have similar acuity, when it comes to non-linear distortions, this is not the case. Importantly, training helps a ton. I am able to pass transparency ABX tests that 99% of the people cannot. Training lowers masking thresholds substantially.
Beyond trained listeners, I have also run into individuals that not only matched my listening abilities, but exceeded it! Someone younger for example should be able to do much better than me.
So be careful in running with graphs in textbooks. Read books like Zwicker and Fastl and see how their listening tests are done (usually a handful of people -- likely university students).
To be sure, most audiophiles are bad at hearing non-linear distortions so if you want to go by that, you can. But please don't generalize if you are not trained yourself or don't know about training.
The above also applies heavily to some distortion tests of speakers and headphones. One was mentioned in Andrew's video which was a collaboration between Sean Olive and Listen Inc. That result simply is not informative. You have to use a) two headphones with identical frequency response and b) including trained listeners who are able to hear non-linear distortions. Otherwise you have ad-hoc results.
Tests of audio compression in research always include this type of trained listeners for the above reasons. We don't want to put our head in the sand and think everyone is as bad as the general public and leave audible distortions in the codec.
3. Minor thing but the historical context is quite wrong. SINAD is an old term but was never used in the manner we are. Here is the Wiki definition of it:
As you see, it is a power metric and is strictly used in the context of communication/RF. We don't use power in SINAD or THD+N.
SINAD's use as equivalent to THD+N (inverted and expressed as dB) is an invention or extension made by Audio Precision in its measurement software. And this use has only been the case in the newer AP software so it has only been around 10 or so years, not a century as implied in the article and video.
It has become famous and gotten the meaning due to my use of it, happy or sorry to say.
When I first started to measure DACs, I only looked at THD+N. Quickly I lost track of what DAC produced what. I turned on SINAD in the dashboard and immediately liked that it was in dB and bigger numbers meant better unlike THD+N which is backward and fully of decimal places. In order to keep track of things, I started to make a bar graph out of them which has not grown to incredible number of products.
Conclusions
The premise of the article/video is wrong. It promotes SINAD to audibility metric and then complains with limited view that it is not. Well shoot, ask us and we would have told you the same thing. To declare it useless and something to be avoided would take a hell of a lot more than what is in the article and video.