I will not buy anything that requires a connection to the mothership unless I have no choice and I really need it. I suppose the phone itself is an exception.
As to the valve requiring the manufacturer's server existing in the middle of the transaction, well, I don't think that's necessary at all. One can certainly connect to devices inside the network through other means.
Of course, in some cases one has to be technically capable in order to accomplish this, I'll grant you that much. The average consumer isn't equipped to manage this sort of thing. And that's how, in the name of convenience, they give up privacy and personal data.
Not sure there's a canonical solution for this. My thermostat should not need a connection to the mothership to work. It should be able to live entirely within my home network and no data should ever escape it. Same is true of my solar array. This is one of the reasons for which I designed, bought and installed my own system. Everyone wanted a connection to the mothership. That's just nuts.
From the standpoint of consumer protection --in terms of privacy, security and product longevity-- I think all products should function under a "privacy-first" policy. What if the manufacturer goes out of business? The consumer's investment evaporates? Same thing in the case of a company being acquired. The acquiring company could very easily discontinue the service. Five, ten, fifteen years from now? Absolutely.
Having devices tied to servers just to be able to function is not, in my opinion, a service to consumers. It's a service to the company making the device. Everyone is selling data now.
I might also go this far with regards to wearable electronics or IoT. They almost have to have a consumer advisory warning, like cigarettes: "This device may become inoperable any time in the future as it depends on a connection to an external service for its operation". I guess I can say that, under certain circumstances, if the consumer if fully informed, no problem.
I'll give you a personal example. Many years ago (iPhone 3 era) we authored about a dozen educational apps for kids. This was around the time that apps were on a slippery slope towards zero dollars. We maintained these apps out of just loving what they were doing for kids around the world. However, at some point, Apple demanded that they be updated to the latest and greatest. It made no financial sense at all do stay on that crazy rollercoaster. So, we made the only decision that made sense: We stopped updates. Within a year they were all gone from the App Store.
Imagine that, but with hardware devices from a range of companies who, after many years, either have to move on to their support their latest products or simply go out of business due to realities of life. This is not good for consumers.
Again, I will admit that in some cases this is a really tough call. However, there's difference between a product made by Apple and one put out by a startup with barely a year under their belt and a runway that might evaporate six months later if sales don't go well.
I would say: Fight hard to make a product that can exist on its own. Make that a selling point rather than a deficiency. You can have enhancements that require getting outside the network, but think of the consumer. Put a different way: If a product cost $1000 and it turned into a brick if the company evaporated. How is that good for consumers?