I think you're reading the article from the wrong perspective. The author isn't speaking to the people using these features, he's speaking to the people making these features. From a user's perspective, you are absolutely correct. People can simply ignore features. In fact, many people do. From a developer's perspective, that's the problem: many people are not using the feature that you made. If people aren't using the feature, then it was a waste of time to implement it. The post is providing guidance towards making features that actually get used, and were therefore not a waste of time.