My hobby is metal machining. Precision is a matter of understanding the physics of what you're doing, knowing your machines, and, well, the machines.
I can give you close to 0.0001 tolerances with the machine in my apartment (about 6-7 times better than stock cheap mills that haven't been properly tuned, depending on the material). Use a big, rigid, production-oriented machining center built for it, and you can do better.
I can't build you a car, though. Precision has nothing to do with this.
> But Apple could easily hire 10-20 people from car industry to fill the missing gap and get there in 5-6 years.
Something tells me you haven't been part of a project that involved mass production, complex engineering (not software-engineering; engineering), or regulatory compliance.
If I were to guess, I'd say Apple has already hired at least 10-20 automotive industry lawyers, just to get moving on just the regulatory side of it.
Any auto made in the last 20 years is incredibly complex. The biggest advancement has been hiding all that complexity.
I have swapped engines and welded cars together, so this is coming from experience. Every car has many sensors, sometimes more than 100. All these sensors feed into enormously complex software state machines written to do all calculations on a RTOS in bounded time and in a verifiably correct manner. This complexity started in, but doesn't stop at, the engine.
You are vastly underestimating the effort needed to duplicate a modern car. There have been times where I've sat for several minutes just marveling at the complexity and beauty of small parts I can hold in my hand. Some of them were surely a years work for an engineer. On any car there are thousands of these clever little parts. For roughly 20k you're buying a product of many thousands of years worth of work.
What we've seen so far from the inside of various ECUs is that they are definitely not always operating in a 'verifiably correct manner', in fact, it's hard to say for sure if the mess of code has had any review or QA imposed on it at all. You'd think so but then it would have looked a bit better.
Electric cars would/could be vastly simpler. Simpler if they don't over engineer the vechicles. The current electric vechicles do nothing for me. I hate vechicles that I can't work on.
My hope is Apple builds a simple, open source, except the assisted driving, vechicle. A vechicle that a weekend mechanic could work on? Something like a 80's Toyota. Just a people mover, that lasts.
My fear is they will build a mac book pro on wheels. I've been told the last thing a buyer cares about is the ability/ease of repair. I just don't get it. I don't like things I can't easily repair.
There will be the day when the good job goes bye bye. Plus, there's a freedom knowing you have a chance at repairing something youself when that warranty goes bye bye?
I would love Apple again if they built a practical electric car. Put a few bells, and whistles in it, but don't it impossible to repair.