His SICP coauthor Hal Abelson also cowrote Turtle Geometry, which explores stuff like topology and the basics of general relativity in Logo. An underappreciated old book I'd like to see more like.

To be fair, when you actually look at Structure and Interpretation of Classical Mechanics (which is a great book), it's still math that a physicist recognizes, not so much an ALGOL-esque program. The Scheme system was extended to handle the math rather than the math brought to the language.

What they did was make all the math explicit, mostly by inserting all the necessary pieces of differential geometry that were glossed over in all other treatments.