Someone once said that laziness is what kept #Haskell pure, and that's the actually relevant feature. This makes we wonder: will theorem proving languages like #lean, where logical consistency is what keeps them pure, deliver the same elegant experience, while avoiding some downsides of laziness (complex runtime, complicated performance characteristics)?