Could Facebook really scale a massive group video-chat service to all (or even half) of its users in that timeframe? This might be advantage an (initially) smaller socially network has over a big one - highly resource-intensive things like videochat and real-time video watching are more feasible.