Why do prediction markets tend to beat polls, panels and expert forecasts?

Because a pundit who's wrong loses nothing, and a trader who's wrong loses money.

Every number on the board is a forecast someone put capital behind. When you look at a market you're not reading opinions — you're reading how much people are willing to lose to defend one.

That's also the honest limit of it: markets are only as good as the money in them. A thin market with $400 of volume is one confident person's opinion wearing a probability costume.