Sure, this is absolutely true. That said, if you can test a prototype for $100 and validate the business idea it's easier to figure out how to pay for (and/or that it is worth spending the time building yourself) the better, built from scratch version. You'll probably also get some good feedback.

I think this is the disconnect for me when reading this question. If the question is "where do I find good ideas for my personal tinkering projects" that's one thing, but if you're looking for profitable startup ideas it's not about the engineering - it's about testing quickly and validating the business so the development piece can be much less important (unless of course the business is software tooling)