It has! Search for "2020-12-01 (3.34.0)" here: https://www.sqlite.org/draft/changes.html
The 3.35 release added more good stuff for CTEs, as well as a RETURNING clause and a much more flexible UPSERT.
The 3.35 release added more good stuff for CTEs, as well as a RETURNING clause and a much more flexible UPSERT.
I think it's still the case that SQLite doesn't permit the use of recursive CTEs in subqueries.
I have make a copy of the Collatz Conjecture CTE that you linked to and was going to see if I could get it to work in SQLite for the next release cycle. I don't (yet) see any reason why it shouldn't work to have the recursive reference down inside a subquery, as long as there is only one recursive reference. No promises. We'll see how it goes.
Have you thought about offering an interface with an intentionally-Turing-incomplete subset of SQL in SQLite, for making queries that can be guaranteed to terminate? Because it doesn't sit right with me that SQL is Turing-complete now. https://news.ycombinator.com/item?id=26529789 goes into more detail on decidable query languages that can still accommodate transitive closure.