No one actually knows how to lculate what speaker cable they need

For the purpose of damping the membrane speaker cable doesn't matter, since it is a fraction of an ohm in series to many ohms of the speaker impedance (breaking back EMF current circuit).  Best effective DF you can get is about 1.5 and DF>25 (not to make it much worse) is OK. Benchmark calculated that for the purpose of audible voltage drop, assuming load change from 3 to 18 ohms, DF=250 is sufficient.  The problem here is, that you might not want amplifier to do that (forcing current at impedance dips).  That's why perhaps many like low DF tube amps.

Let me examine one statement made before by b4icu:

The DF is actually the way the Amp. is getting control over the speaker's coil. A coil, especially a moving one in a magnetic field, generates an electric current that is equal to the one made it move, but in the opposite direction. It is called the "Lentz" law.
Speaker cables are in a way two resistors (Rc) that connect the power Amp. to the speaker. Why two? becuse the cables go to the speaker (red) and goes back (black) to the amplifier.
As so, the speaker's impedace has no significance in this electrical circuit.

 Of course it has big significance.  Speaker's resistive impedance (equal to about 2/3 of nominal impedance) is in the circuit.  The whole idea of damping comes from the fact that voltage created on the speaker terminals (known as the "Back EMF"), by the motion of the coil in magnetic field, produces current (flowing thru speaker) that causes opposite membrane motion effectively stopping the membrane.  This current flows from one speaker's terminal thru speaker wire, amplifier's output, another speaker wire, another terminal and the coil.  Changing resistance in the circuit from 0.1ohm to 0.001ohm means change in total circuit resistance from about 6.1ohm to 6.001ohm (for 8 ohm speaker) resulting in very small change in the "braking" current.  

I don't know why do they make very thick cables.  Perhaps to reduce inductance (straight wire inductance is slightly lower for thicker wires), but this would not make sense since the same can be achieved by better twisting of thinner wires (that reduces inductance and increases capacitance).    One possible reason can be skin effect, that starts at gauge 18 (20kHz, copper).  I'm not sure if it is audible, but remember Audioquest FAQ.  Stranding wires even with isolated strands won't help since wires are still in magnetic field of each other.  Placing wires in helical twist on hollow tube serves two purposes.  It subjects strands mostly to magnetic fields of neighboring wires only and provides twist.  This twist not only reduces wire inductance but also provides immunity to external electric and magnetic fields.  Since each of perhaps 20 or so strands has decent gauge (for the ease of working with) it makes overall gauge much lower.  That's how my Acoustic Zen Satori is constructed.

One observation  - amplifier's output has very low impedance for low frequencies only.   High frequency electrical noise induced in the cable is injected into the output of the amp and amplified by the amp because of negative feedback.

As for cable directionality - yes, AC charge flows forth and back but energy to speaker is delivered one direction only, on the outside of the cable where other factors (including insulation) might play role.  Again, I don't know if direction is audible or not - but it is much more complicated than it seems.

Plain zip cord is fairly inductive and thicker zip cord, that is the stuff with larger gauge copper, becomes even more inductive while the resistance drops.

Inductance of straight wire is lower for thicker wires.  For instance gauge 16 wire has inductance of 0.33uH/ft  while gauge 10 wire inductance is 0.29uH/ft

as a copper wire alone shall have no inductive or capacitance at all!

Of course straight wire has inductance. It is on average in order of 0.4uH/ft that represents about 0.2ohm at 10kHz for 8' wire.

unfairlane, claiming that only speaker cable or amp's output impedance affects damping, without taking into account resistance of the coil, shows lack of understanding of basic principles.  It also appears that you share this lack of understanding, but at least I made you smile.

b4icu, yes there is a lot of ignorance in this area, so let me try to explain how membrane damping works. When amplifier outputs positive voltage membrane goes forward. Motion of the membrane is caused by electric current thru the coil. When membrane moves on its own in the same direction it produces back EMF voltage of the same polarity. This voltage will produce current that flows from the speaker to amplifier. This current has opposite direction now, producing force that moves membrane in opposite direction - in effect stopping it. This current is equal to back EMF voltage divided by impedance in the circuit including cable, amp’s output and speaker itself. Even if we assume only resistive part of speaker’s impedance (bass frequencies) it is still about 2/3 of rated impedance. For an 8 ohm speaker it will be likely around 6 ohms. So now you have 6 ohms in the circuit and likely another 0.1 ohm of the woofer’s xover coil. Reducing down cable’s resistance (0.05 ohm for 2x10ft gauge 14 wire), won’t change anything since you already have 6.1 ohm in the circuit.
So, for the purpose of the membrane damping best effective DF=1.5 (nominal 8 ohm divided by resistive part of the impedance). When we choose an amp with DF=25 it will lower overall damping by about 5% (in comparison to DF=2500). Changing wire gauge from 16 to 4 will improve damping by 1% only. The easiest exercise you can do is to take woofer, short it and try to move membrane by hand. It will be hard to move it, regardless if you short it with short jumper or long cable (you won’t feel any difference).