It does not really show what the practical advantages of those gold capacitors are for audio applications in particular, though. Of course, a good design does not have many electrolytic capacitors in the audio signal path in the first place, regardless of whether they are gold or not.
Here and here there are some samples recorded from a Xonar D1 - which has a lowly CS4398 DAC, NJM5532 op amps, and generic non-gold capacitors - that can be compared against the original audio. Chances are that you would have difficulty hearing the difference without knowing which file is which.
Yes, the implementation is everything, ie the sum of the parts rather than just the parts. I haven't finished my critical analyzis of the SQ of the Soundblaster Z, but as it looks right now so does the Xonar D1/DX indeed sound better than Soundblaster Z for pure stereo listening. As I said earlier the Soundblaster Z has a slight coloration to the sound, most easily detected in the treble. None of the sort to point my finger on detected from the Xonar DX. But the Soundblaster Z has other advantages, as better gaming support.
The difference between the Xonar DX and the more expensive Xonar ST is a closer call, definetely no real verdict yet, but it feels better to use the Xonar ST. Not very objective