I don't think the MIT guys have the same motivations as the author of this book. He (Walck) discusses the suitability of (a subset of) Haskell in this article:

I don't think the MIT guys have the same motivations as the author of this book. He (Walck) discusses the suitability of (a subset of) Haskell in this article: https://arxiv.org/abs/1412.4880 https://arxiv.org/abs/1412.4880 Maybe someone else can shed light on the MIT mindset. Certainly some of Walck's points apply to Scheme as much as to Haskell, but Scheme lacks the type system, syntax and syntactical "convenience" of curried functions. The basic strength of functional programming is the lack of complex imperative book-keeping: your code looks more like math. My impression is that SICP and SICM are eccentric.