If you say "people" want something else than what is offered now, then by providing this something else you should be able to capture a lot of the market. Example: Apple produced an iphone, but some people wanted more open, not-so-closed-garden, customizable, moddable phone. And they wanted an OS to go with it. And here it it, Android and a booming Android phone market. Yes, a giant - Google - owns the OS. But another giant - which are many around - could dislodge them if they slept at the wheel.
Another example: Microsoft owned the browser market with their Explorer. Then they fell asleep at the wheel and now Chrome owns the market. Sure, again, it took Google to do it, but it can be done. And it will be done again - when there is a substantial unsatisfied market need.
If this is not happening for a long time, then I think it's time to consider the possibility that it's not that the "market" "fails", but just your ideas about what people should want are not the same as what people actually want. If nobody buys the "ideal" phone somebody built - maybe it's not actually ideal for enough people.
> If you say "people" want something else than what is offered now, then by providing this something else you should be able to capture a lot of the market.
Suppose you could capture a lot of the market in the sense that the installed base of people using your phones goes from 25% to 40%, but then your repeat customer interval goes from an average of 3 years to 9 years. Your future sales outlook would be down. You're not going to do it.
Now suppose that your existing installed base is below 1%, or you're a new company and it's zero, and by doing this you could capture even 3% of the market. Then you do have the incentive to do it, as long as that percentage of the market can recover your R&D for developing a competitive phone to begin with. But what happens if the existing market is vertically integrated, so you have to do a lot of unrelated work from scratch instead of starting with a competitive reference design? Then your R&D costs are much higher and you don't do it.
And then nobody is making the competitive phone that people want.
> If nobody buys the "ideal" phone somebody built - maybe it's not actually ideal for enough people.
Let's pick some random midrange device; Samsung A35. It's ~$265, has 8 cores at >2GHz, 8GB RAM, 128GB storage, 2k display, etc. All I want is that with a removable battery and the ability to run a mainline Linux kernel. These are similar specs to the Fairphone 5, which doesn't yet appear to run a mainline linux kernel (they're working on it), but that's $800. Even at that price they're selling them, but which phone is the open competitor in the A35's price range? Can you identify anyone offering this device?
You can't say the market doesn't want something the market isn't offered.
Another example: Microsoft owned the browser market with their Explorer. Then they fell asleep at the wheel and now Chrome owns the market. Sure, again, it took Google to do it, but it can be done. And it will be done again - when there is a substantial unsatisfied market need.
If this is not happening for a long time, then I think it's time to consider the possibility that it's not that the "market" "fails", but just your ideas about what people should want are not the same as what people actually want. If nobody buys the "ideal" phone somebody built - maybe it's not actually ideal for enough people.