If anything, this has made me rethink possibly going over to common lisp, which I was thinking of for the larger stdlib, because SBCL is very fast, supports TCO, and you can disable case insensitivity.
Common LISP was made to smooth over the differences of the ancient LISP's plus standardize their situation a bit. If those don't matter to you, then the alternatives are often better. SBCL being one of them. If I get back into LISP, I think the Racket Scheme community looks like it's delivering most bang for buck in terms of what the tools can do after investing time learning them.
Most schemes don't have a lot going for them in terms of libraries, which is the primary reason I'd want to go over to Common. I really don't like racket, for mostly aesthetic reasons. It seems a bit over complicated, trying to stuff a lot of cutting edge research into its core, and forcing you to learn about 6 different models at once just to understand what its documentation is saying. Its OO system is apparently lackluster, especially since most schemes fork tinyCLOS for OO, and syntax-case is too complicated for its benefits.
I'll stick with CHICKEN for now. Its module system may be a bit awkward at times, (you can't cleanly bind all your dependent modules into one executable), but it's much more minimal, and while its documentation may not be the best, at least I can understand what its saying.
If anything, this has made me rethink possibly going over to common lisp, which I was thinking of for the larger stdlib, because SBCL is very fast, supports TCO, and you can disable case insensitivity.