Successful startups are on a thin line where the product is just compelling enough to be a stand alone app and then going deep on satisfying the needs of the app's main premise.
There must first be a deep need, solving a fundamental problem people are hacking with other apps. Then an entrant introduces a new and compelling single action, a desired feature, a new feeling, and builds from there. These are not "single-feature apps". They have many features and will have many more. The difference is the features are built specifically to make the users experience around the apps main premise faster, simpler, and more fun than current solutions… not to optimize for features.
Apps who can serve user needs in this way are finding themselves in very powerful positions. The best tend to define their vertical and quickly build an ecosystem around the app's main action. And because they are solving deep, sometimes untapped, problems (of quality, networking, productivity, etc) they are out performing established companies in the space. Users are completing more actions in less time, leading to greater usage/trust/growth/community.
I expect this type of efficiency to eventually be present in every aspect of my wired interactions. It is a race to user-centric utilitarianism.