Forgot to add, in order for any jitter to be present for 16-bit (not factoring in jitter reduction by the DAC nor dithering), it needs to be less than 346psec ((1000000000000/44100)/(2^16) ).
So, since <346psec is litterally not present for 16-bit, without reduction or dither, making the claim that going from 22psec to 7psec was drastically audible should now be clearly evident as bogus (placebo).
And yes, there is no benefit from going higher than 16-bit, which can cover say 35dB to 126dB without dither, and 25dB to 145dB with noise shaped dither.
So, since <346psec is litterally not present for 16-bit, without reduction or dither, making the claim that going from 22psec to 7psec was drastically audible should now be clearly evident as bogus (placebo).
And yes, there is no benefit from going higher than 16-bit, which can cover say 35dB to 126dB without dither, and 25dB to 145dB with noise shaped dither.