That advice mostly comes from the fact that most startups don't have any semblance of a business model. So, the only thing to tweak is the product in search of blowing up user "engagement" numbers.

But even with a business model, the product needs to evolve based on customer engagement and feedback, which sometimes takes us to a place we would have never imagined in our vision.

I don't think I understand your stance on when to put your visionary foot down and when to listen to customers.