So they fully admit that the iPhone "cannibalized iPod sales" but then said that they "didn't count the iPhone because the product is also a wireless handset." So all this is saying is that among mp3 devices sold that weren't also phones in the week of Aug 23 - 30, Sony had 43% and Apple had 42.1%, with no mention of what the +/- error was in the survey. While this might be an indication of a trend, it is also very likely to be statistically insignificant. But that doesn't make as interesting a news story.