Quartz has a thought-provoking article on The hardest trick in mobile design: making the product simpler. As an example, it shows how Foursquare started with something simple, made it more complex over time and then went back to having something simple.
I see something similar when working with companies. There’s sometimes a tug-of-war going on between departments and/or people that means there ends up, over time, too much in an app. This is to satisfy all the interested parties. People who think they know best, sometimes don’t. Features sometimes don’t get used and, if the company is forward thinking enough to have analytics, such features get removed. Of course, all this isn’t new and was the case on the desktop and web.
An observation I have is that the ‘featureworthyness’ of an app is often related to how well the inventor of the app knew the actual domain. By domain, I mean the specialist area of the app (retail, automotive, health, farming or whatever). Those closest to the problem tend to know what’s needed. They have also probably used related apps in their specialised area and already know what works and what doesn’t. So, find and use someone with a deep knowledge of the domain.