That would be true if you could show that candidate "types" are actually meaningful distinctions and not just the founder's first impression based on limited data. I think that's a stretch.