How is GP missing the point? They're arguing about the timescale of entropy, not the basic principle. You're right that ordered systems eventually become disordered, but "eventually" can mean 10^100 years, and you haven't provided any evidence to the contrary.