SCAMmy Co-recursion
The continuation to which the destructor is applied is turned into a single application (with the destructor pattern-matched against those in the continuation). The knot is tied by replacing any instance of the co-induction marker (•) in the resultant term, with an abstraction re-inserting the continuation.
...really, world? Is anyone going to care?