The problem I see with this heuristic is that it is going to give false positives for, well, the vast majority of cases where there's a supposedly new idea that actually is just an unnecessary rehash or combination of things that already exist but with a new coat of paint. I don't see how the heuristic can possibly have any predictive power. Look at pretty much every tech startup, every get-rich-quick scheme, every cryptocurrency project, etc. and they'll be making grand claims that their idea is groundbreaking, and technical people are going to say "couldn't you just x?" or "isn't this worse than the existing x?" and those dismissals are actually going to be reasonable and proven valid in the overwhelming majority of cases. Of course, your heuristic will also catch the tiny number of cases that do turn out to be groundbreaking (or at least extremely successful or popular).