Excel workflows are terrible though. No version control, hard to test, prone to indexing errors. And doing very sophisticated things with it gets hard; lots of financial analysts/quants are moving over to Python for analysis anyway.

If you’re thinking about this in isolation, I can see why it would seem a bad idea to move power-excel users to Python. But take this in the context of a much wider shift where many shops are already shifting to Python for other reasons, and so we need a way to help transition the Excel power users over too.

Excel has its place for sure, but I think it’s interesting to consider whether another tool paradigm could gradually replace it; we would need to really hone the flexibility and expressivity of the UI for simple tasks. The benefit would be that when your task grows you don’t need to re-implement it in a new Python engine.