The book has about 4 paragraphs on the subject, I will summarize:
He starts out by saying that speaker cables are usually misterminated. In particular, misterminated at high frequencies where reflections would matter most. (I don't know that I quite believe this or understand what he is trying to tell us).
He goes on to say that most speaker cables have a characteristic impedance Zo between 50-150ohms. (I think this is the heart of his opinion) TL effects are unlikely to matter much in the freqs in the audio band, but they can cause load impedance variations that can provoke amplifier instability at high freqs.
He goes on to do some math on a 10ft cable and calculates 18Mhz as a freq that could be problematic, above audio amps freq range, but can cause amp output oscillations (if no L-R isolation on the output).
He shows some impedance plots for open, short, and terminated connections, it is obvious that there isn't a problem if cable is properly terminated. He also states that a Zobel network at the far end of the cable is a good idea (even if only 100ohm R and 0.01uF C).
There you have it. I think this all seems pretty reasonable, but I will defer to Gene, et al, for now.
<!--[if gte mso 9]><![endif]--> It seems pretty reasonable from what your saying.
Basically any properly designed audio amp is severely bandwidth limited at RF frequencies.
Most power amps have a small inductor in series with the output that will normally block any RF from getting in as well. Power amps have a low impedance at audio frequencies, but not for RF because it's well outside the normal bandwidth and can't be mitigated by feedback. Some amps have a small cap (~22pF or so) across the F/B resistor so RF can theoretically get straight back to the input stage where it's demodulated and comes back out at audio frequencies as buzz (TV frame buzz), or sometimes speech/music. That usually needs a lot of RF though!
Also, a Zobel network helps (a bit), but most use a 10 ohm resistor and 100nF cap, and that doesn't match the cable impedance of 'normal' cables (normally around 100 ohms, give or take). It is probably a good idea for speaker manufacturers to add a Zobel in the box as part of the crossover, or ideally right at the speaker terminals.
In order to get a speaker cable with a characteristic impedance of 8 ohms [sqrt(L/C)] would require the cable to have an ungodly amount of capacitance which in itself could cause an amplifier to oscillate. Goertz makes such a cable by sandwiching two flat conductors on top of each other and they put a zobel at the end of it to stabilize the amplifier b/c the capacitance is like 6nf/ft!
I did an article on such a cable long ago here should you care to read it:
Speaker Cable Reviews - Faceoff 2 | Audioholics
<!--[if gte mso 9]><![endif]-->
In further speaker with Emotiva, Ray added the following:
ANY cable that is not driven and terminated with an impedance that matches the impedance of the cable will likely ring. This has nothing to do with RF pickup from the outside world so shielding is not the solution in this case. The shielding does change the characteristic impedance of the cable which means it may ring at a different frequency or level but it will still ring unless the impedance of the cable matches the output and the load. The thing is that no practical audio amplifier is generating high enough slew rates to cause appreciable ringing and what ringing there might be cannot be reproduced by the speakers, nor could you hear it if it could be reproduced.
You see practical application of matched impedances every day with the 75 ohm cable distributing the cable or satellite signal to your DVR. Ringing on these wires would cause extreme degradation of the RF signal. Not because it magically get back into the source device, but because the reflected signals cancel the incoming signal within the cable. The down side of matched impedance is that 50% of the power is consumed in the output termination and 50% in the load. Not a huge deal when you are sending milliwatts of cable signal. Huge deal for an audio power amplifier! <!--[if gte mso 9]><![endif]--><!--[if gte mso 10]> <![endif]-->
<!--[if gte mso 9]><![endif]--><!--[if gte mso 10]> <![endif]-->
