I decided to save cartesian products and power sets for a possible
future combinatorics SRFI, as Alex Shinn suggested.
+1
Those combinatorial algorithms have a lot in common with lazy sequences, as addressed by e.g. SRFIs 40 and 41. Intermingling lazy sequences with eager data structures such as these SRFI 113 sets and bags is a challenge, though. I suppose we should take this one step at a time.