[rapjr9]

I noted the same thing 15 years ago and asked around about it. A prof at the college I worked at said tools like this (PCB design tools are similarly awful) don't have a big market, so there is not a lot of competition or funds for development. Maybe that is less true today with more people designing their own PCB boards and 3D printed objects. There is also lock-in, if all your CAD is in AutoCAD are you going to scrap it all and start with a new tool? While there is often some ability to export to a common format and import to a new tool, often lots of little important things get left out. So the tools don't get improved much because the vendors don't have to improve them since few people have a choice, and even if there is more of a customer base today the barrier to entry is high since the tools are complicated. The free tools also make financing improvements difficult since they reduce profit. Open source and open standards can help somewhat here, but things like the odd interface Blender used for so long can limit adoption. BRL-CAD which started life as a US Army project is interesting:

https://brlcad.org/

I haven't used it lately but it was very stable, reliable, and powerful. It had a command line interface which made it somewhat clunky to use, but it was a very capable program. I think a GUI or two have been added to it to make it easier to use. It uses constructive solid geometry (adding and subtracting solids) to create objects, though I see they have added B-Reps to it.

The fact that the early CAD tools all used command line interfaces instead of GUI's also may explain some of the awfulness, many of them tacked a GUI on top of the command line interface instead of rewriting the software from the ground up, and some of that funky command line weirdness may persist in the strange workarounds the GUI's had to use.