KLowe, first of all there are no such things as subharmonics. That's a holdover from 17th century music theory. They were a fantasy all along and they simply don't exist. Difference tones, sure, but subharmonics, no.
Second, transients don't put any special strain on a digital recording system; electrons just don't have enough mass or inertia. If a system's frequency response covers the audible range and if it can handle a full-scale sine wave at the top of that range without clipping or spewing crap all over the spectrum from slew rate limiting, then it can handle the audible part of any transient you care to throw at it. Anything with a rise time faster than the signal I just described is putting out energy above the Nyquist limit by definition--so yes, when you retrieve that signal it won't look the way it did originally, but there's no reason for it not to sound identical.
That's not even a digital issue; it's just as true for analog tape and vinyl LPs, which have high-frequency limits, too. You can't cut a 2 kHz square wave on an LP that will look anything like a square wave when you play it back, but so what? Music still sounds right on a good system and wrong on a crummy one. That's as true for 16/44.1 digital as it is for analog tape, records, or radio. There are examples of purely analog systems that have audible and measurable problems in this area, and there are examples of digital systems that have no audible or measurable problems in this area.
Lots of people seem to get confused about transients. It's important for the audio circuitry to do either (or both) of two things: (1) cleanly filter out any signal components that are so high in frequency that we can't hear them, and/or (2) follow the original signal through all its rapid twists and turns. In terms of human audibility, either approach is precisely as good as the other, and they can be combined nicely, too (e.g. a circuit preceded by a clean, simple low-order low-pass filter at, say, 40 kHz can also be designed with a moderately high slew rate--but then it doesn't need to be extremely high, since the filter guarantees that nothing much above 40 kHz will ever have to go through it).
> The 16-bit/44.1kHz technology used for Redbook Standard (commercial) CDs cannot capture or convey enough audio detail in the transients, harmonics and sub-harmonics, the ambiance (room noise, air noise) etc., that creates the full, warm, sweet sound we are used to hearing in a live situation. That is why many people complain of CD recordings that sound harsh or brittle or missing the "room" sound. Recording at 24-bit/96kHz allows us to capture a world of sounds in the 10kHz-20kHz range that include all those extra transients
I can't agree with a single thing in that whole remarkable statement, sorry.
--best regards