Hmm... you're quite right, and it turns out the Guardian's own site has an incorrect link in it, which I didn't spot.
What's happened in terms of ranking is that Birmingham City has shot up in computing due to a huge surge in overall student satisfaction (from the 60s to 90%). In one year. Surprising, but not entirely unknown.
I looked up the current and old unistats data on the two courses, as that's often enlightening.
Aston still wins overall I think - some key data that is important:
Aston is BCS accredited; Birmingham City is not. While British Computer Society membership is not vital for everyone, if a course is not accredited, it means it does not meet the criteria for practical skills required for BCS accreditation.
In terms of employment, 36% of Birmingham City graduates were unemployed at six months; while only 15% at Aston were (it was a bad year, so figures are generally high). In other words, Birmingham City has more than double the unemployment rate.
If you look at what jobs those who did get work do, the picture is similar: 95% of working Aston graduates having a professional or managerial role, while only 65% of Birmingham City graduates are (to save you the number crunching, this means that in the end Aston has nearly double the chance of getting you a professional role at the end of the course.
I noticed, interestingly, that the average actual fees paid at Aston are lower than at Birmingham.
There is a change between 2011 and 2012 data (which are the last two years) for Birmingham City - which is a major upswing. However, there can be (in both directions) quite marked differences between years. Averaging out over the last while, Aston is markedly ahead, but last year on a some questions, Birmingham City did outperform them.
Overall, I think Aston look the better bet from the KIS data, with better consistency of student satisfaction and, importantly, markedly better employment. The Guardian table this year seems to be rating "value added" higher, but that looks at grades on graduation, not employment. The former is generally easier for an institution to adjust than the latter.
I hope that helps,
George