If the product is ready to ship, then ship/market it and see how it does. If it's not ready to ship, then do what research you can to make sure it's not absolutely stupid and then build it and find out. At the end of the day the market is unpredictable past the broad strokes.
Sorry to shit all over your idea, something that would help validate prior to shipment would be nice. I'm not even morally opposed to the idea of pitching people a fake product and seeing if they click, you'd just need a way to separate the fake product from the real thing you're building and hope they come back for the real deal, and that could be tricky.