It is your right to believe what you asserted. My experience points me in a different direction; my comments are not directed as criticism, just offering a different perspective. See my comment below that I'm well aware there are personalities for whom the perspective I offer are not suited for.
> Prices act as a mechanism to incorporate information on efficiency of resource use and cheap disposable items are cheap in part because they take fewer resources to manufacture and transport.
I consider the concept of pricing-conveying-information outdated in a world that moves much faster and is far more sophisticated and complicated than earlier ages. The modern drawback with relying upon pricing to convey information is there is a huge latency between pricing and customer-relevant information coming out about the product or service. This latency only gets larger as software-aided and -enhanced customers start grappling with an ever-larger constellation of externalities of the products and services they incorporate into their decisions, as do sellers.
I don't want to find out the BuyItForLife tool I purchased for a premium price is not the same quality as a past generation of buyers bought and recommended, but only after half a decade of market pressure forces the manufacturer to lower prices as the news of their shoddier modern instances start to gradually filter through the public. I want to know before I buy it, and it is the Internet, not pricing, that is gradually forcing out that information asymmetry that exists between sellers and consumers in category after category. It is not pricing chiefly disambiguating the dozens of choices facing Amazon buyers, it is the recommendation data.
Furthermore, repairability and durability in tandem are often consonant with modifiability. The Internet enables hundreds of thousands to millions of people to access and share with each other skills to modify all sorts of products, far faster than before. The barriers to entry to learn product hacking skills like soldering, welding, plastics molding, woodworking, etc., are far lower than someone born just 20 years ago can imagine, and we aren't nearly finished yet with that trend. Product modifiability is essential to not just keeping products out of the landfill, but also future-proofing and adaptability to a user's specific requirements. I hold that such adaptability is essential to confront the ever-complex future we all face.
There is most assuredly a spectrum of trade-offs. In a resource-constrained setting (perhaps below about a middle-class First World standard of living, maybe lower?), I most heartily agree with you that over-engineering every product choice is premature optimization. But the "particular use case" mindset, even when the use case is say 20 years, is a subtle trap. That mindset is a much-scaled-up timescale version of the "why the poor stay poor" time-capital preference decisions. The farther out you can push that timescale, the more you (and your progeny) have an opportunity to escape that trap.
Note I'm not saying buy over-engineered for every single product, even if you can afford that purchase behavior. For specific categories of products however, like objects that are intimately tied to physical human factors, it's a good bet to buy over-engineered where you can afford it and/or it has a positive impact on your productivity. For example, there is unlikely to be a quantum leap in wood planer technology akin to the transition from hand-held to power planers in the near'ish (20-100 years) future; a hand-held power planer you buy today can conceivably be handed down to a few future generations and still deliver the same functionality. Products that can enjoy a quantum leap in functionality from embedded software could be a big exception in the coming decades, though; change from a hand-held to a bench planer, and software could be a significant consideration in the future, for example.
I don't advocate this purchase behavior for everyone. It works best if you have the personality who considers in advance the logistical tail of properly maintaining a purchased product, including proper stowage (often means climate-controlled or at least humidity-controlled), preventative maintenance, ancillary repair tools, and spare parts stocking (a lamentable side-effect of a minority purchase behavior is spare parts are not as readily obtainable in more remote locations, so stocking the most commonly wear and tear parts mitigates downtime). My rough rule of thumb is 1.5-2.0X purchase price for that logistical tail for the time that I own it before passing it onwards. If you don't plan on having children or a similarly close relationship with relatives from a younger generation (nephews/nieces for example), or mentees, then this purchase behavior applies less to you. And so on. Different strokes for different folks, there are no hard and fast rules here, just considerations to weigh.
Finally, we all face the minimal quantums of time and energy in our lives. These products represent embedded time and energy not just to create them, but the buyer's time and energy, to evaluate, understand operation, adopt to idiosyncrasies, etc. Product switching costs these days are increasingly incorporating greater complexities. For an ever-increasing swath of products as software permeates across the physical landscape, it is no longer just a consideration of keeping embedded time and energy from a landfill.