I was involved in the design of an ASIC. This was going to be used by people from all different areas of the company, and they all had various demands. Most had never worked on a chip before, and didn't realize the cost of what they were asking for. The easy thing to do would have been to ensure that every requested feature was added But that would have burned area, power, and huge amounts of design and verification time.
Luckily we had a senior guy who had the clout to say "no". Over and over again, he would tell people how much things would cost in terms of money and time, and ask if the features they wanted were worth that much. Or he'd point out that what they wanted could be done with firmware plus existing features. Overall, he probably saved millions of dollars and caused the project to be delivered on time.