Personally, if Apple creates a standard car simply with a great software experience (I'd expect at least a few small novelties), and not too expensive, then I am totally in.
a) You can operate it without talking to it
b) It has dedicated knobs/levers/switches for primary essential controls.
c) It doesn't require internet connectivity, continual tracking/telemetry etc.
Apple knows how to build interfaces. Including interfaces outside of software. I suspect it'll be the primary thing they focus on in the product launch video, how they didn't screw up the UI and there will be knurled titanium wiper speed ring switchy things and tactile feedback in the damn mushroom leather wrapped heated steering wheel and Jony Ive whispering directions, seemingly inside of your head.
I don't want Apple forcing their excellent design sense down my throat, I'd much rather have decent design I don't have to use.
I do assume it'll be a locked-down, walled-garden experience, only serviceable at Apple Stores (which will open garages at some locations), assuming they can get away with that legally. They'll reject any non-Apple-genuine part, of course.
But to your point about voice control, I do expect Apple will try to make that really good. Touchscreens suck for keeping your attention on the road when you want to change the heat temperature or change songs or stations. Physical controls are better. But good, painless, non-frustrating voice control is by far the best option for avoiding distractions and keeping your eyes on the road.
Just one tiny example: see the manual dial or “crown” on Watch, AirPods Max, and now Vision Pro. You could easily imagine those products without an unnecessary manual “dial” when there is both voice (Siri) and software controls. But, you still have a “dedicated knob”.
Not sure which iteration of it you have, but I've actually found the software on my 2022 E-Class to be surprisingly decent. I generally prefer plugging in my phone and using Android Auto, but if I don't need Google Maps to get where I'm going, I usually don't bother with it and the car's own system is fine.
You ever heard about Autosar as a software middleware that is in use for years by most of the OEM to decouple software and hardware development?
For example, Daimler/Mercedes is developing for years toward self driving cars. They had working prototypes in the 1970s with "electronics" based on analog designs that was working as good as a modern Tesla FSD, with the exception that it required lots of space. Mercedes/Daimler is the reason why we have cheap ultrasonic sensors in modern cars. They developed that in the 1990s together with then Temic. They had a working self-driving prototype during that time. All that lead to cheap sensors we use today daily for adaptive cruise control in every modern car by any OEM. Mitsubishi had for sometime the best in class safety system with sensors able to detect pedestrians, like kids, and bicyclists.
MobilEye is now developing self-driving capabilities for years, and IRC first used by BMW for safety features in their models. Only later Tesla used the system in ways which were never consumer ready. That'S why they finally split.
The things is OEMs are conservative introducing features to consumer. Because nobody wants to risk its brand to devalue. Its not only that, if there are too many accidents that would mean that the insurance will have an extra premium, which would lead that consumers don't buy such cars.
So I was driven in car in the early 2000s with features that are rolled out just recently, like the level 3 by Mercedes on highways.
The things is, most just don't see how much things are going on under the hood. They just see the MMI/entertainment system. Yes, that was for long not as high value proposition.
With the physical parts of the car commoditized to oblivion, the primary opportunity for product differentiation is in the sensors, software, and UX. All the automotive OEMs know this. For better or worse, the traditional automotive companies are terrible at this part of the business and they do it reluctantly. It isn't in their DNA, and their traditional production processes are unsuited to it.
When the two big execution problems of building a modern automotive company are "software, sensor, and UX design" and "efficiently managing complex global supply chains", it is easy to see why Apple might be uniquely positioned to be successful at it.