[berkes]

> about trusting your own feelings about a project

was also when I started disagreeing. Sure, your gut-feeling is a good guidence. Asking yourself "do i need this polished" or "would I stop using the software if X is removed" is great advice.

I disagree because I think better advice is to make it slightly more scientific: "Most people would jump ship to a competitor if I remove X" is easily proven (just remove it), but also researchable: remove it for a cohort. Ask 10 random customers. Check your event-log or database if X is actually used currently.