FYI attached a spreadsheet (Excel, in zip file) showing the voltage loss and amplitude variation for various cables assuming a perfect amplifier (ideal voltage source). User inputs are in green boxes for cable length, nominal power, and speaker impedance nominal, minimum, and maximum values. Note power is used to calculate voltage loss just for fun but has no bearing on the relative amplitude at the speaker terminals (drops out of the equations). The spread in dB tells how much the SPL might vary at different impedances, which might occur at high or low frequencies and thus change the speaker's frequency response. This is the main reason IMO speaker cable lengths at home might matter; nothing to do with propagation delay, it is about changing the frequency response. Low resistance matters more with low-impedance speakers and ones with low impedance dips, natch. You could also use this to compare two different lengths (or gauges) of wire to the same speaker.
Not responsible for typos/mistakes in the equations -- I am more into Mathcad, Matlab, etc. than spreadsheets for this sort of thing but wanted to us a program most people might have on hand. If you find an error, please let me know and I'll fix it if I can.
HTH - Don
Examples:
10' cables to nom 8-ohm speaker with 4-ohm min and 16-ohm max
20' cables to 4-ohm speakers with 1.5-ohm min, 12-ohm max: