Disclosure of lifetime costs needed but many pricing options valid
i don't have a problem with a company designing a product to be easily repaired or designing it to be impossible to repair, each design point has it's advantages. What I do think should happen is device repair-ability and costs before purchase should be DISCLOSED.
Take the John Deer tractor case. If JD wants to sell tractors at a lower initial price, because they believe they will make 30% of the initial sale price on repairs over the lifetime, I'm ok with that, provided the purchaser understands the initial price is discounted based on expected repair revenue. This is exactly the inkjet printer model. The initial prices of many consumer inkjet printers is less than the real cost, because they expect to make money on the ink cartridges. You also can buy commercial inkjet printers, for rather higher initial prices, that use huge bottles of ink. Epson for example seems to sell with both sales models. Both ways are valid, but you should understand the trade-off.
I don't think manufacturers should be forced to sell at disposable or manufacturer only repair initial prices, and then be required to make it cheap or easy for later repair. I'm personally a fan of DIY hardware repair/tinkering, but I also do engineering for a living, and know a lot of markets are initial selling price sensitive, often more than total cost of ownership sensitive, so shifting some of the revenue later in the life-cycle allows more sales. If you buy a $200K tractor for $100K, don't grumble that you're spending $5K/year on repairs for the next 20 years. I could also see dual prices on tractors, pay $200K now and have free access to repair data and guarantees on part availability and price, or pay $100K now and depend on manufacturer service forever, perhaps with a no-lemon guarantee/lifetime repair price cap.
Another one of the the tractor arguments is perhaps tractors have just become TOO feature filled, and if they stripped off the stuff that was not essential it might only cost $100K instead of $200K. Sophisticated farmers argue that extra $100K improves their productivity more than $100K over the life of the tractor, and other farmers, who perhaps are not willing to adapt their process don't find the extra features. This raises the question: should we have "extravagant feature" laws. i did some calculations an pure electric tractor, and it's initial price would need to be like 5x as high, in return though your energy costs might be half as much as diesel fuel. If you could keep that tractor busy 24x7, like if it were self driving, you would have a tractor that had a lower cost per unit of work done. This is exactly the electric self-driving Uber/Lyft model. Often a more expensive machine has reduced operating costs.
As far as I can tell, many modern $250 phones do most things almost as well as $1000 phones, so the extra $750 you spend is just a status symbol thing. Phones are not things you can share, so a $1000 phone does not cost less to operate than a $250 phone. The price dynamics of tractors and phones seem quite different, so am not so sure the same strategy and laws should apply.
Another option in this whole problem would just to not sell things, only lease them. So you could not buy a $200K tractor, you could only lease a $1000/month tractor, that included all maintenance+parts, and a service up-time guarantee, essentially TaaS (Tractor as a Service). If you did this, manufacturers would be very motivated to minimize lifetime maintenance costs and maximize up-time.