My gut feeling is that if you have a group of committed potential customers -- i.e., people who trust you and are already interested in what you're building --

My gut feeling is that if you have a group of committed potential customers -- i.e., people who trust you and are already interested in what you're building -- then it will probably work. But I'd be wary before you try to get people you barely know to do your beta testing for you unless your product is extremely sexy. The buy-in has to be there before someone will commit to helping you improve a broken product.