OpenSCAD is often said to be for programmers, but in some ways it's more for mathematicians. If you know the formula/code for the object, you can calculate the bounding box, but that isn't a way most programmers are used to thinking.
I can calculate the bounding box, or indeed the exact position and orientation of some surface I want to add to, but that’s tedious and so I don’t want to.
It’s like programming without higher level language constructs. Yes, lower level languages are still Turing-complete so can technically do everything. But the necessary housekeeping distracts from the real problem I’m trying to solve.
