To your first point, of course people "talk their book." It's a chicken and egg problem. And though it's in theory possible for someone to convince the market that the price of something should be higher, it's virtually inconceivable for this to happen in the global energy market, especially if the party in question's financial interests are publicly known. (In less liquid markets, this can become a valid concern.)

Your second point is much more solid. I'd like to see more empirical data regarding people's risk preference across different contexts and magnitudes. Seems like an interesting topic.