Trouble with this is that the feedback loop will still try to null out the difference between the input signal and the fed back signal from the speakers which are now at the end of the feedback cable. So, in taking out the effects of the speaker cables, you introduce the effect of the feedback cable. Strongly suspect that either method makes cock all difference. Buy decent gear and all this cable fetishism is for naught anyhow.

More interesting was the Philips Motion Feedback idea which took feedback off a sensor on the speaker cones. It was, unfortunately, implemented by Philips.