This is an important point that often seems lost when abstracting out concepts like "power" and "control."
There is some theoretical sense in which a user of free software has a lot of control, including the power to change it. Software being free can have a lot of indirect benefits to the user, but ultimately practical "power" is a function of what you can practically do with it, and that is very different from what the thing can theoretically do.
Apple probably has motives for the app store other than making user experiences simple. But, they also really have the goal (and the effect) of making the user experience easier, which puts more real power into real hands. If the below median user can install a photo sharing app to see her nieces photo collage, that's real power in users hands.
Also, I think being strict and abstract about "exclusively to advance the interests of its owner" is losing a lot of important texture. First, very few technological benefits go exclusively to anyone. They go to users, vendors, various other users and vendors. A lot of them get released into the ether as externality vapors, like productivity or ideas for new technology.
The reality of the current state of computing is interesting, but it's very hairy and hard to grasp and boil down to simple abstractions. When we use our phones, we are interacting with other people, generating data/content on other people's phones, in various databases to the benefit of a bunch of people/companies. When you are using your navigation app, generating data about traffic, you are benefiting the app, other osiers. Possibly future users. Possibly that data will be sued in other ways. Maybe a construction company will buys it to plan new roads (and win a contract) which will benefit the city… Maybe someone's girlfriend will one day have it beamed to her implant to estimate when Brian will arrive.
Is this bad because it's not to the exclusive benefit of the user? The benefit to the user would not even be possible with this mesh of stuff.