A few of my thoughts on cables, noise, dynamic range, and sound quality.
1) Many subjective descriptors of sound quality, e.g. 'space', 'transparency', 'haze' et al have their root in noise. Noise or noise products within the audio spectrum. RF noise, above 100KHz or so is so far above human hearing as to be totally irrelevant, unless noise products falling within the audio band can be shown to be produced via intermodulation or other plausible mechanism.
2) A DAC not performing a flawless conversion it is by definition defective.
3) Above the Nyquist frequency, sample rate, or X-Axis value is progressively less important. By comparison, bit sample depth, or Y-Axis value must be at least 4 and preferably 8 bits greater than desired dynamic range to ensure resolution of low-level signals. In reality, 96K sample rate at 24 bits resolution exceeds the capability of even the best the human ear by a significant degree.
4) Reconstruction filters do matter, as sound quality changes are repeatably demonstrated, and are highly dependent on how the upstream ADC functions. Therefore, changing reconstruction filters should be plentiful and easily chosen. The phenomenon known as 'digititis' has as one of its root causes progressive use of suboptimal reconstruction filters producing audibly poor output. The other major cause are compression artifacts from e.g. low resolution MP3 files.
5) Claims of lower noise do not stand up to scrutiny of the system dynamic range as measured from typical room noise floor (~40dBA) to typical system peak output capability of 110-115 dBA (70-75 dB total dynamic range). Reducing noise from say -100 to -110 dB in a system with a 75 dB dynamic range falls into the category of 'measurable, not meaningful'.
6) System gain structure can only be optimally set using a signal generator and oscilloscope. For example setting the preamp to clip 3dB above power amp maximum peak output. A misconfigured gain structure can easily cost 20dB or more of dynamic range.
7) Network cables and switches simply cannot impact sound quality. The very nature of the protocols used ensures bit perfect transmission. Otherwise, the entirety of the internet could not and would not function.
8) Analog cables may potentially impact sound quality due to interactions of cable capacitance and inductance with downstream input impedance.
9) It occurs to the casual observer that using cables as very expensive tone controls is an interesting behavior.
10) The most significant variable in every system is room acoustics and room noise floor. Until those issues are addressed one is spending dollars to pick up dimes.
