I've actually been thinking about this a lot lately. I belive psyco is just a special case of partial application.
nfib is faster than fib because it's compiled with the knowledge that its argument is an integer. But you could make it even faster if you knew that its argument was an integer that fit into a machine word. (That might not be what you want to do with nfib, but I imagine a lot of numerical functions are going to be used almost always with numbers that fit in machine words.)
My thought is that you should be able to specialize functions for arbitrary predicates. For instance,
(def fib (n) ; yeah, it's a bad definition
(specialize (and (number? n) (< n (exp 2 32)) (> n 0)))
(if (< n 2)
n
(* n (fib (- n 1)))))
Or maybe you could even declare specializations outside of the function. For instance, imagine writing
(specialize (fib n) (and (number? n) (> n 0) (< n (exp 2 32))))
right next to some code that really needed a fast fibonacci function. And of course, if you think about it like that, there's no need to restrict it to function calls - that's just an arbitrary restriction. Let's say you need really fast mapping of a function.
