I agree very much with the idea of "improving on something that already exists". I hate it when you tell someone an idea, and they rule it out because "it already exists". There is always a better way to execute on the same idea.
However, I am not a fan of "Has business model other than advertising". Although it is something to have in the back of your mind, it is hard to say what the product/idea will evolve into when it actually becomes something that can be profitable.
I think a careful evaluation should be spent in finding the right balance between "General enough, so you can find users" (as advised in the article) and "if you're a solo entrepreneur find a niche so you have less competition, expecially from big names" (paraphrasing Rob Walling, Tim Ferris and others)
Follow this strategy to a spectacular fail. We've all seen the startup kids who invest 2 years+ of their life into a startup they really believe in, THEN they release. and it flops.
Your belief in your startup is useless, other than as a way of keeping you motivated until you ship.
This is why MVP and pivot and "throw-jello-at-the-wall-and-see-which-sticks" exist as better, adaptive, strategies.
I agree. But even by following the MVP methodology we still need to pick an idea to start with. Regardless the belief that execution is the key and idea worth nothing, I would still pick the one which satisfies most of the points.
However, I am not a fan of "Has business model other than advertising". Although it is something to have in the back of your mind, it is hard to say what the product/idea will evolve into when it actually becomes something that can be profitable.