Rupert Neve shares his stories at Blue Rock Studios 

I go back and forth on this a bit, myself. Certainly high frequencies do interfere with lower frequencies to create new interference tones. However, those should also be captured if they're within the audible range. So is that why it sounds good? Probably not. It would seem that Rupert is saying that having the actual "inaudible" frequencies there is important to perception of the music as real. He might be right. It could also be due to losing those interactions in mixing. If you haven't captured all those frequencies when you record instruments in isolation, they will not interact with other instruments captured in isolation, but you'd have the interactions that were actually recorded within the isolated track. However, there's another possibility, here. That is, that by designing the equipment to handle high frequencies like that, the quality of the parts has to be higher, and they're probably going to operating within a very comfortable range of their design limits. It's kind of like when you capture higher frequencies in digital. You can use a much gentler filter to low pass up to 192khz than you can for 44.1khz, for example, because a gentle filter that cuts off everything above 44.1khz will have effects down in the range that you actually DO want to capture and reproduce (up to 20khz). Those are analog filters that are being used to do that as well, on an ADC. So if you have designed your gear to reproduce up to a lower frequency (say 20hkz), it's more likely that there might be consequences of that in the audible range. In the end, though, RND gear sounds amazing, so I dunno how much I care if the philosophy that produced it is correct for the reasons stated. It worked.