Yes, that's assuming optimal gain structure, I admit I didn't fully read yet, but there are definitely use cases where huge digital attenuation followed by high gain will deliver audible hiss, when you have 100 steps of digital volume and you find that it's already too loud at 10%, can't be good, The problem is not the digital volume so much, if you have no other way to control it's a problem, and yes, the main thing is that sensitivity of earphones/headphones are litterally all over the place. Analog volume control are useful gain switches too, sometimes it's the right thing to do.
edit: yes, after reading, the statement is true, even 50 dB attenuation beats the dynamic range of the best amps, but the problem is than there are plenty of situations that you can only use a small fraction of that dynamic range. Bottom line, we all know the conclusion, the basic Idea is this, you'll never attenuate the noise floor with digital attenuation, it will stay fixed, you can just bring the content closer to it.