The most obvious omission from that model is that in a real economy people voluntarily give money to others in exchange for goods and services. If everyone was equally good at producing useful goods and services then even if the yard-sale effect was occurring due to investment/wagering behaviour it's unlikely to lead to the extreme inequality the "pure" version does. It's also fairly obvious that in the real economy there's virtually nobody in the super-rich list that's got there purely by being lucky with investments. To what degree that's true only because we do have redistributive taxation systems I don't know.