I've just realised that I have confused this paper with another paper, which I think was from the racket people and involved what they called "nano-passes". I t

I've just realised that I have confused this paper with another paper, which I think was from the racket people and involved what they called "nano-passes". I think I've actually glossed over this paper many times for this reasons. I'm not interested in writing a scheme compiler (there are several good ones out there already), but this looks like it has a lot of useful techniques I can pilfer.