There are a few industries that actually use this model, with strange results. (National talent hunts that consider unknown singers and actors come to mind.) Most industries have moved beyond the give-me-more-names approach, for good reason.
Screening candidates is crucial to good hiring — and that is hard work. Nobody does it with 100% effectiveness. The world is overflowing with brilliant people who got dinged quickly in broad hiring searches … to the ever-lasting regret of the hiring organization. Facebook turned down Brian Acton in 2009, when he wanted a solo engineering job, only to pay $22 billion a few years later to bring him on board via the purchase of WhatsApp (which Acton and Jan Koum built on their own when no one would hire them.) Harvard Business School turned down Warren Buffett way back when, because it didn’t see the future billionaire as being a worthy addition to the class. And so on.
Suggesting that companies consider 500 people pretty much obligates screeners to make rapid decisions on the basis of very limited information. That’s a recipe for hasty (and often foolish) decisions. Do it that way, and the best candidates are all too likely to end up in the discard pile.
If you really want to help companies hire better, come up with better screening methods. Nobody is perfect at going from 50 to 8 … from eight to four … from four to one. Everyone would like to get better. There’s a huge amount of lost value to be captured at every stage of the candidate-assessment process, especially in the later rounds. Shoveling a lot more names into a broken funnel doesn’t really meet the needs of hiring managers, HR, growing companies … or the talent itself.