Lowest cost possible is often fun optimization problem engineers enjoy. Save a fraction of cent here and there and there too. And in some ways it is good for consumer getting cheaper products.
On the other end you have something like Juicero. Massively and wastefully overengineered piece of crap. To do not that useful task. While being extremely expensive. And probably not actually last that long.
Maybe one day if far future we end up with some mature balance between two. But I doubt it...
I don't think that engineers think "okay, so if I use this chip, the product will last 4 years, so I can use this other chip that will last only 2 years because it's a few cents cheaper".
If you want it to last longer, it's a lot of work: you have to somehow test the components you buy (or get those who produce them to do it) and then you have to test whatever you build with them. So you have to invest in it, it's not just a design decision.
Same for waterproofness: it's not that you actively drill holes in your device to make sure that it won't be waterproof. It's just that if you actually want it waterproof, you have to design for it, then you have to test, and iterate a few times. If your consumers still buy your device if it's not waterproof, then there is no need to invest in waterproofness. But it's not "planned un-waterproofness".
On the other end you have something like Juicero. Massively and wastefully overengineered piece of crap. To do not that useful task. While being extremely expensive. And probably not actually last that long.
Maybe one day if far future we end up with some mature balance between two. But I doubt it...