There are even ways to "parameterize" control point driven primaries, allowing them to be driven by points, lines, and planes. This can give teams the "best of both worlds", should it become a sticking point.

Trust me when I say blends and such can be built fully parametric. I have done complete door inner's and IP's that were parametric. Really complex shapes that passed inspection the first go around and such. On many occasions. I have done bumpers and fascias as well but as most people know doors and ip's have very complex shapes.

The adjustable nature of the models were such that if the arm rest grew pockets moved, the entire door outer that mated to the sealing surfaces as well as IP to door gaps were controlled parametrically. Which was good because when new door swig studies were done and clearances were determined all I had to do was swap one surface out and dozens of blends and mating surfaces updated in minutes. Not hours.

Again some surfaces were parametric some were not. It all depends on requirements of the final design.

That is where people get to hung up on parametric everything. Sometimes you have to but others are not necessary. Mostly, primary slabs are control point based. After that mostly parametric design. It works. Trust me I have done it for years. In UNGX and in Catia V5. Not just once on a fluke but dozens of times.

I would love to show the work but I can not. That whole getting sued thing...