I don't think the GP's solution is really that unreasonable - it's just that there's something missing. Like you said:
> Modern goods are fractally complex: individuals can't be expected to become experts in all the disciplines required to judge the quality of a good.
I think the idea that individual consumers can individually exhaustively assess the value of their goods is not really reasonable.
However, I think it's reasonable to believe you can construct a system where a large number of consumers, enthusiasts, and experts pool their knowledge together in order to much more accurately assess product quality.
For instance, if we could conduct wide-scale long-term surveys of the reliability of large appliances, e.g. by requiring that appliance technicians log service visits for an appliance into a database (that could be run by a non-profit), then that would significantly increase pressure on manufacturers to build robust products. After all, if the cheap $2k fridge is listed in the appliance database as having a lifetime one-third of that of the $5k fridge nearby, far more consumers will opt to buy the latter, because now they actually know that there's an advantage to buying that one.
Also, consumer testing labs like Consumer Reports and ConsumerLabs do some expert testing of products, which is better than nothing.
I think that the problem is information asymmetry - companies know how bad the products that they're selling are, and consumers don't. The free market system only works if buyers have "good enough" information to make a rational decision (although that's a necessary, not a sufficient, condition), and we're not in that state right now.
> Government interference in purchase decisions is a big hammer. It is hard to imagine policies that would actually help.
I agree - it's far harder for the government to legislate vague things like "appliances must be built well" then something like "appliance lifetime and service visits must be logged into a registry".