Fact #1. higher sample rate needs better clock, low phase noise, low drift & ultralow jitter under <10ps.
Fact #2. Highs over 10KHz suck with 44.1Khz.
becouse the Nyquist theorem its wrong,
the diference between a sine waveform, square or saw waveforms its the harmonics, sines dont have harmoincs.
any square/saw signal over 10KHz needs harmonics that go over 22050Hz, if those harmonics are gone, becouse the jitter or becouse the Nyquist limit of 44.1KHz those Square or Saw become distorted sines that look like unstable triangles.
at 44.1kJHz, all the highs sound "the same" distorted sinewaves & brighter becouse the harmonic foldback, like a harmonic exiter included for free at 44.1KHz.
but at 44.1KHz theres no diference between harmonic shapes in highs.
everything sounds plastic & unreal over 9KHz at 44.1KHz.
Fact #3. better external clock needs better shielded cables & clean stable power.
Fact #4. SADLY!!!
if you EQ at 44.1KHz, and press the 96KHz or 88.2KHz button,
or...
if you EQ at 96Khz and press the 44.1KHz button or 88.2Khz or 48Khz,
most digital algorithms inside plugins dont work the same at diferent sample rates with same settings, becouse the harmonic and jitter diferences between sample rates, and you will have to re-do everything again.
thats why some people think that 88.2KHz or 96khz does not sound better, but it does, and a lot better for highs, the mids sound more crystal becouse thers no harmonic foldback from the highs.
at 44.1Khz sounds dirty.
Fact #5. the SRC algorithms, most suck,
http://src.infinitewave.ca/ SRC Comparisons Fact #6. most internal clocks suck.
Fact #7. some DACs sound better at double sample rate.
becouse the anti-aliasing filter design.
Fact #8. You need twice the CPU or DSP at X2 Sample Rates..
Fact #9. Latency gets half with same buffer size at double sample rate.
"converters work faster"
Fact #10. working at any sample rate with a bad clock sounds worse = darker and blurr.