LP is basically the idea behind JN, yes. Though it's a bit limited in that sense; for example literate programming often allows arbitrary ordering of code snippets and extracting parts into separate code blocks (like `<<foo>>` in the linked post).

But Jupyter Notebooks does it more like a REPL where every code block can be executed separately, which is nice too. Afaik this is also possible in Emacs' org mode.

Mathematica had notebook interface since 1988, and Maple from 1992. You'd think these were the examples that Jupyter notebooks are copied from.

That’s actually a really good comparison. Jupiter adds interactivity that WEB didn’t do, but remove the ability to easily compile and distribute the finished product.

Another extremely important successor to Knuth’s ideas is the built-in documentation comment formats that most programming languages have. They tend to eschew linear narrative in favour of navigable hypertext, which is honestly a good idea for many purposes but sadly relegates the “overall vision” stuff to supplementary documentation.

I think LP sounds more like SCM (particularly the git log). You interleave your thoughts (commit messages) with the changes to the code (diffs).

Or rather Jupyter notebooks are a tool you can use to do LP.