Speaking as someone who has actually designed sample rate converters (SRC) for ADCs and - more importantly in this context - actually listened to the effects of sample rate conversion, this video and the assertion that SRC is so good you can't hear the difference is just plain wrong. To make SRC work completely transparently is almost impossible, and to get it to work tolerably well requires incredible precision and oversample rates - something that crude interpolators inside computers are incapable of doing. And this is when I am SRC from 104MHz down to 98MHz (for eventual 48kHz decimation). Going from 44.1k up to 192k is very much more difficult, as you have the transient timing issue to worry about too - and the interpolation filters for SRC ignore the transient timing reconstruction issue completely.
When using my DACs do not use SRC or upsampling, always ensure bit perfect data is given to the DAC - then the WTA filter will do a very much better job at reconstructing transient timing than any SRC or external filter. Remember that the WTA filter is the only filter that has been designed (with thousands of listening test to optimise it) to reconstruct the original timing of transients as accurately as possible. Transient timing is absolutely vital for perception, as the brain uses the timing of transients to perceive instruments as separate entities, place them in 3D space, perceive LF pitch, determine timbre, and of course the stopping and starting of notes - vital for rhythm. Music without timbre, pitch, instrument separation and rhythm is just nasty noise.