Does this imply that lambda expressions in Haskell and ML don't have a "coherent design around scope and extent"? This is quite a claim, to be honest....