So when you combine this with the Standish study that said only 20% of the features on custom developed software are used always or often, you have a pretty good argument for simplification. Not only do all these extra features never get used, they cost a whole lot more to implement than the basic features that are needed.
So next time you're in front of the customer and they ask if you can do this or that, say yes, but it will cost them. Ask them what the real business value is. If they have a good answer, the feature probably makes sense. Otherwise, focus on the top 20% of the features they want (based on business value) and deliver that first before committing to any additional work.