Thanks for posting this. It took me down a rabbit hole. I saw KFW perform ~generators in SF in 2011 or so... and it blew my mind. Funny, it took me another eight years to get into physical modular synths. And I'm only realizing now how impactful that performance was on me; he's improvising lines, like a jazz solo played on a clock divider or driving an entire system from simple gestures. What's amazing to see is his comfort around chaos and how well he navigates to disparate plateaus of sound, one after another.
KFW makes it clear in this video[1] that he's using a combination of a few modules based around a concept or abstraction (e.g. chaos/feedback in clocking via OP's link). It's so refreshing, and this approach appeals to me because you get to hear the boundaries of the sonic character of his voice(s).
I think if you build a system around his approach, you'd be well served by complex voices like those from Industrial Music Electronics (he mentioned Zorlon's Cannon), Schlappi Engineering, many others. Importantly, would be to pick a "shape" (e.g. clock feedback) that integrates chaos/feedback and develop gestures. I've been working with XAOC Drezno for sometime and use it to transform simple CV into arabesque pitch and gate information that drive an entire generative system.
Something that's brilliant about KFW's work is how he utilizes control that feeds-back with audio. It's something that I've been working with recently and can be rewarding but challenging to contextualize. I think this approach may lend itself better for compositions that use on a single line/single voice.
Curious to see what you come up with. Please post a system, as it happens.
[1] Resonant