Cool idea - is the % of foursquare users in a sample population consistent? What if Tuesday is "bus in the foreigners/elderly?" - valuable info regardless, but I'd be curious to know how that holds up.