Wow...questions, questions ;)
Ok, I will simplify things to keep it under control so bear in mind that it is more complex than I am making it sound.
I am doing some work right now involving modeling of cables and their impact with various loads (mostly inductive ones like motors and speakers). All cables have what is called the "characteristic impedance" which is a function of the series parasitic inductance (L) and parallel parasitic capacitance (C). The former is mostly wire-material dependant and the latter is dielectric/sheath-material dependant.
The speed is the "signal propagation speed" which theoretically is the speed of light but (also) due to the L and C of the cable, it is lower in actuality. The speed is inversely proportional to the product of inductance and capacitance - in other words, the higher the inductance and capacitance, the slower the signal will travel through the cable. This obviously impacts dynamics.
Now, here is the fine line that causes confusion. The frequency issues due to these parameters are mostly in the MHz to GHz regime which is of course not audible by humans. However, harmonics are produced by electrical interactions between the L and C which will produce sidebands that theoretically extend ad infinitum in both directions which means they reach the 20-20kHz region as well. Whether or not they are large enough in signal magnitude to be heard is still under hot debate everywhere. My personal opinion is that in some freaky way, we are able to hear this effect which accounts for the differences in sound between various cables and equipment.
The frequency response really doesnt register these effects because it is a time-domain function. If you look at it in the frequency domain (via the often mentioned Fourier transform), you can see the harmonics involved in the signal. BUT, current measurement techniques dont really have good enough resolution to get everything so part of the information is missing.
This leads to my infinite variable dictum which goes something like this: everything affects everything all the time. All we can see is our resolution capability very similar to our stereos. The information is infinite in quantity and so the amount we experience is a factor of this resolution. What happens outside the resolution is unknown. What we do know is that the effects of parasitic L and C of cables (and dont forget all the wires in your components too) add up to something that falls within human ear resolution.
Anyway, I blab about this forever so if you have any questions, feel free, but dont forget that no one knows everything that goes on in our systems. Arthur