Who in the world decided to use such a stupid method for their graph? "Average number of weeks between software purchases"? Please.......
Also, there is one critical flaw I see in this whole attempt to make the PS3 and Wii look like software selling phenoms... and that's multiple games/software sold at the same moment. It seems more likely to me that Xbox 360 gamers have historically bought games in large packages. They purchase 2-3 at one time, thus why they go longer in between before buying more.
Just something to think about before believing this very deceiving graph. How else could you explain the 360's much higher overall games per console sold, if they were buying games less often? It's simply because 360 gamers tend to purchase multiple titles at once.