There was some discussion on that recently.viewtopic.php?f=5&t=68669&p=2022519

oops! sorry! thanks for the heads up

phew! that is quite a long walk. but if anyone finds this one instead, TL;DR: you use it for sustain/length.

i watched this clip on youtube the other day, where Steve is mixing a record and wants more thump out of the kick. so he sends the kick via bus to a parametric EQ (GML, i believe). he boosts some low frequency content and returns this to another channel. he also sends some of that return channel back to the GML for some feedback, but not to the point of oscillation. it's fancy and sounds cool, but why the hell would you choose to do this instead of just using an EQ either on the kick itself or even a parallel channel?