I'd take this a step further and say that syntactically-significant case, a la Haskell, is a good idea. You already use case to disambiguate to human readers, so why should the compiler be oblivious to the distinction?