I think the best way to evaluate an idea is to ask if you'd use it yourself. If it solves a problem for you, there will be others that want it - nobody is unique. It's when you think you know how to solve someone else's problems that things go wrong - collaboration is important in that case.