How can you commit to a fixed amount to execute on an idea that requires an unknown number of iterations, re-evaluation of initial assumptions, and customer feedback? Maybe this is possible IF your customer is a sensible designer who can iterate in parallel with you and has a realistic sense of what an MVP is and isn't. If your customer is just some schmuck with 2.5 grand to throw around and no product experience, you'll likely find yourself fighting time consuming battles of scope and deliverables. And how can you juggle multiple projects like that simultaneously, when there's unavoidable friction that comes with the context switching? And, how fulfilled will you feel when your day-to-day is filled building things that are bug-laden and just minimally good enough? What kind of portfolio will that generate for you?