This concept is inherently flawed. These complex devices were built by teams of highly skilled and specialized engineers. The concept that you should be able to repair it with a screwdriver and a 6-pack is outdated.
I grew up on a cattle ranch, and we had lots of tractors and equipment -- some of it over 100 years old. I rebuilt an International Farmall Model B Tractor with my Dad when I was 14. And it was a lot of fun to go around to old junk yards and find the parts we needed. This tractor was built in the 1940s and I think that's what people think of when they think tractors.
I learned how clean an maintain simple engines, how to organize parts so I could confidently take something apart and put it back together again, how to bend and weld metal. It was a great project. Totally 100% think every kid should do something like this. When the tractor was done, and repainted, I got to drive it in a parade.
When we were putting that tractor back together... a lot of times bolts needed to be cut, drilled out, holes re-tooled, or just welded back into place. And while it looked fine, there's no way someone should think we were professionals. I'd be scared if someone took that tractor and tried to do real work with it.
There was one place in particular, where I used a bunch of bailing wire to hold a non-visible part where it needed to be held. And it worked OK for driving in a parade. But if someone had attached something to the PTO shaft... and it needed significant torque, I'm guessing the bailing wire MacGyver patch job would have caused catastrophic failure.
And that's fine, if it happens as soon as you do it, but like, what if it happened while you were working in the field, or what if someone else (20+ years later) starts the tractor up and uses it -- not knowing that my splice is in place? It makes the machine fundamentally unreliable, unsafe.
Now... that was an 80 years old tractor. Fast forward to today, when the machines have conservatively 1,000 times more power. Ranchers cut corners all the time... can't wait a week for parts? Just splice something. But how does the software know to account for that? It's expecting a load of 1,000 pounds and it adjusts hydraulics accordingly... but what if due to the "fix" it can now only support a load of 300 pounds?
And let's say someone is hurt? Is that on John Deere? Or the Rancher who bought the tractor 2nd hand at an auction? Is that on the insurance company, that has to insure equipment that costs 10s if not 100s of thousands of dollars that get plugged into the tractor and rely on it operating to spec? It's just such a massive can of worms.
So let's then add the software element. And assume that some fixes are going to not just be hardware, but software-related as well. Would that mean that John Deere, or Apple, or anyone, would have to make their code open-source -- for all of their competitors to review as well? Or provide peer review if you want to make a change? For how long? Like let's assume tractors today last as long as the tractor I rebuilt with my Dad in the 90s... 50 years? Wonder how you even factor in 1/2 a century of support in to the purchase cost. Oof.
Anyway, it's just a mess. People over simplify it, "I should be able to fix my bike, so I should be able to fix this $500,000 tractor..." and at some point it just doesn't scale.
And, like others have said... I really don't want to have to run my phone through some sort of unit testing / compliance check every time I go to get on an airplane. Batteries, and phone software... they're a whole other complex issue. Even what we have now, with 3rd Party batteries... it's pretty risky.