Power Cord Lenght ? Better shorter or longer?

For all audio applications, shorter is better. For IC's, speaker cables, digital cables and power cords. For power cords and speaker cables, inductance is the parameter to be minimized and it increases with length.

If you are using the cable as a tone-control, then none of the above applies. But doing this guarantees that you will never get an accurate reproduction of the original performance.

What I am saying is that all cable designs become more inductive as you lengthen them. The inductance per foot does not change, just the total inductance. Just as all cables add capacitance as they get longer.

Designs that have very low inductance tend to have higher capacitance per foot, due to coupling between the conductors.

My goal is to minimize the inductance in a power cord. I dont believe in filtering-out high-frequencies because this puts impedance in the way of transient currents, which must flow unimpeded. Particularly bad for power amps. Some filtering may make sense for line-level and digital components.

As far a power cords are concerned, I dont pay much attention to the fact that the voltage is a 60 Hz sinusoid. I pay more attention to the di/dt of the current, of which the spectrum can contain very high frequencies. This is what any inductance in the cord will impede.

As for speaker cables, we are definitely on the same page. I believe that extremely small phase shifts in the audio range (2-5 degrees at 20kHz) can be audible. This is why cables with 1 MHz bandpass sound so good. However, I am also proponent of band-limiting the speaker cable to 1 MHz or so. This is to attenuate natural resonances that are in the 10's of MHz. In a really low-loss, high-bandwidth cable, these resonances can sustain themselves for long peiods. I believe the result can become audible with most amps. I must be doing something right, since the reviewers really like my speaker cable.