The touchscreen enables a bad and dangerous ux design.

Physical knobs—where appropriate, like volume—are amazing because one can adjust the setting almost instantly, without having to keep an eye on the touch screen, and without having to develop muscle memory.

A touch screen still pales in comparison to a mouse and keyboard, for example, so this shouldn’t come as a surprise.

That's a good point. Often when I see UX this bad, it's also usually the result of siloed teams or designers working on individual parts without thinking about the whole experience, and it all just gets stitched together at the end.

I can see the console designer being like, "Oh cool, steering wheel designer is out of room, so I get to design cruise control buttons" rather than pushing back with, "This is a really bad idea"

If you've ever worked at a company whose bread and butter is hardware, you know many of them treat software like just some other line item on the BOM: Like a screw or nut that needs to be as cheap as possible and "meet the spec", however insufficiently defined that spec is. I'd be willing to bet that this attitude is what's ultimately behind crappy software in cars. They don't see software as value-add, they see it as a cost that they grudgingly must eat in order to barely compete with the other car companies who also have crappy software. Yet, these companies who are so bad at software, and obviously hate it, keep adding more and more of it into their products!

It was a lot better when you could just ignore the bad software and use your phone for navigation/music. Now you need to use the bad software to operate more and more essential things in the car. What's next? A big screen in front of your face with a 10FPS animation of a steering wheel that you have to drag around with your finger to steer?

I think that's pretty much what user testing is, and it should be done long before releasing a product.