Actually think founders need to look existential questions in the eye, and figure out reasonable answers to them. Yes the questions are better framed in an open-ended non-demeaning way, but asking them will save you from locking yourself up for years building something nobody wants.

Also, building a better version of X is usually less promising than building the first version of Y. Especially when X is pretty well-executed.

Most top startups I see are carving out their own niche: https://topstartups.io/