Second, all your examples so far used structural subtyping (objects with the same fields have the same type) rather than nominal subtyping of C++ (distinct classes have distinct types even if they have the same fields; the subtyping must be declared in the class declaration). For the structural subtyping, upcasts and downcasts can be done mostly automatically. See the OOHaskell paper or the code Hello Oleg, I've glanced over both HList and OOHaskell papers when I considered taking different approaches. Albeit elegant, OOHaskell looked too heavy for my purposes, I don't need mutability, for example. And HList paper left me with two questions. The first one is how much such an encoding costs both in terms of speed and space. And the second one is can I conveniently define a Storable instance for hlists. As I said before, I need all this machinery to parse a great number of serialized nested C structs from a file.
Best regards Dmitry