From Alan Kafton’s web site for his Cable Cooker,
Cable break-in occurs as current flows through the conductors of wiring components. Dielectric stress from voltage differences between conductors also contributes. It takes many hours of in-system use for wiring components to break in, primarily because audio/video signals from normal program material are so low-level.
Consider an interconnect from pre-amplifier to power amplifier. The maximum signal level for full power output of the average power amplifier is 2 volts peak, and the average signal is much less. Typical input impedance of a power amplifier is 10 kohms at the low end for consumer gear; 47 to 100 kohms is typical for a solid-state amplifier, while several hundred kohms impedance is not unusual for a tube power amplifier.
Taking the best-case values from the above, the maximum current seen is 2 Volts/10k ohms, or 200 micro-amperes. This would not be continuous current, because the voltage value is peak, not rms. One can calculate a "use value" from the above equation multiplied by the total time this current flows. Let's call that the Current Time Value (CTV).
Playing an interconnect cable in an audio system for one week (168 hours) of continuous use would expose it to the following CTV: 168 hours x 0.0002 amperes = CTV of 0.0336.
The CABLE COOKER produces signal levels far higher than those seen in normal audio/video system use. The Cooker's sweeping square wave oscillator drives a high-efficiency "H" bridge MOSFET switching circuit. The output signal is a square wave from below 100Hz to above 16 kHz, plus harmonics. Output voltage is 12 volts rms. The measured current flowing through the interconnect is 120 milli-amperes.
Installing an interconnect on the Cooker for one week results in a CTV of 168 hours x 0.12 Amperes = CTV of 20.16. This is a value 600 times greater than under the most ideal audio system conditions. The "stress" on the dielectric is also much higher due to the higher output voltage. Results with the Cooker are typically audible after less than a day.
The same signal also feeds the speaker cable binding posts. The load at the speaker cable inputs draws a continuous 1.88 amperes of current through the wire. With a potential of 12 volts, this is equivalent to a continuous signal level in excess of 22 watts rms. This continuous signal level played through loudspeakers in a home environment would be unbearably loud. Put simply, there is no way for conventional in-system playback burn-in to approach the intensity and efficiency of the Cooker.