There are several interesting things about this graph, one of which is simply the number of page loads a month, right now there are over 1.3 Billion page loads per month and there is no sign of the increase in traffic slowing. Although I am probably responsible for about 1% of those, I still find it a huge number and wonder how they will cope in the future.
The second thing I like is the subtle implicatons about peoples behaviour you can see in the data, you may note for example that there are several dips in the traffic towards the end of a year, most obviously in recent years. My guess that this is the effect of Christmas holidays, when people are forced to go home, and no longer bother to check the news every five minutes as they do when chained to their desks. This of course also tends to imply that much of the traffic is still coming from the western world.
Then there is something I don't really understand unless it is simply a artifact of how they have graphed the data, but it appears that traffic had been increasing rapidly in the days leading up to September 11th 2001, before the attacks on World Trade Centre and the Pentagon. My guess is that they have simply plotted monthly totals which have tended to make it look like there was a rise in traffic pre September-11th, I would imagine that if you plotted daily totals you would see fairly smoothly increasing traffic until September 10th, then traffic would spike dramatically on the 11th before tailing off over the next few weeks, with a few superimposed bumps and wiggles related to events such as the opening of attacks in Afghanistan or new information regarding the attacks becoming available.
I'd love to see the daily totals for this period, I think this could really be useful in determining the most important events in modern history, especially when coupled with information on the most popular stories, though I rather fear that various celebrity scandals would dominate the events.