With regard to the comments after the aside:
The former model has become a “fad” because of technological progress in e.g. logistics and predictive analytics. The wording “send some stuff to you every month you don’t get to pick the stuff but trust us” belies the fact that this is actually a very difficult problem with a very interesting solution. If some instance of this model succeeds it will be because it managed to create a service that improves lives at _essentially zero cognitive cost_. To cultivate and then engage with certain preferences takes some amount of time and energy and to be able to offload those costs to software is, at the very least, interesting.
The latter model is only uninteresting under the assumption that the products will never work in the sense you mentioned (efficiency improvement). I’m not sure there’s any reason to assume that “smart”, “connected” products won’t improve substantially in the relatively near future. I also think there’s plenty of simple tasks that could be automated in a meaningfully useful way. For example, cooking rice.
More generally, building a profitable, sustainable organization of human beings is intrinsically useful and interesting. This last part is less a reply to your post and more a response to a cynical (in my opinion) tone throughout this discussion.