Programmers like to work on problems which are technologically interesting, and they love adding new features. Dealing with technical debt and solving UX/UI issues isn't as much fun, so in a Bazaar model they'll simply not do it. The result is a product which feature-wise is very powerful, but UX-wise an absolute nightmare to use.
But when there's a party genuinely interested in the product as a whole, they can push more Cathedral-like UX-focused development. Have a handful of devs focus on UI stuff for a year or two, and it has suddenly turned into a world-class product. Blender and KiCad have gone through this before, and it seems like Ondsel is pushing something similar for FreeCAD. Let's hope it works!
I can make FreeCAD go, but we still have to manually create dimensioned printsets for our sheet metal shops, rather than just being able to hand off a drawing file. I feel like when we get to the point of just being able to hand off a drawing file, it'll be in a much better place.
This reminds me of iirc KiCAD 5 to KiCAD 6. Overnight it went from some weird clearly-Linux program to become a viable product, an excellent one even.
KiCAD uses FreeCAD on the backend for things.
I’d love to see FreeCAD take the same path!!!
However… when I looked at it last year the “let’s draw a cup” tutorial was so pathetically bad I closed it and went right back to solidworks without a second thought.