The only way to build great products using your gut it to scratch your own itch. This is a completely valid approach that can work really well, but even then, you're better off doing this customer validation thing. You might be the only customer out there. And yeah, web apps are cheap to build compared to chip fab plants, but why spend several months building something and then another 6-24 months pushing it to get traction (the really hard part) if you can find out that there's no market with a few days of work?
Finally, I know quite a few successful entrepreneurs who are huge advocates of the customer development model, so I hardly think that being a great entrepreneur requires you to avoid talking to your potential customers before you start building, and hoping that you're just the 1 in a million genius who really does get it. You're almost certainly not Steve Jobs, my friend. Survivorship bias is the only reason we think we might be...we don't see the millions of failures who thought they could predict the market's desires without talking to them.
Folks unsure of their product/idea tend to not have a great product or are just not good at starting a business. But the people that know a market exists for their idea and just need it implemented are more than likely to succeed.
I'm racking my brain to think of any successful entrepreneurs that had to get validation before they started. Jeff Bezos? No, he had an idea for an online bookstore, quit his job, moved to Seattle and started one. 37Signals? I bet Jason Fried asked for lots of validation. Steve Jobs? No, they built the Apple 1 and started selling it. Henry Ford famously said "If I had asked people what they wanted, they would have said faster horses."
I just don't get why you wouldn't try to find out if you'll be able to find customers before you spend all the time and money to build the thing.