The big advantage MS tags had in the testing I’ve done is that they are scalable much easier than qr codes. The tags can cope with significant image degredation and remain readable, while the error detection and correction in QR codes is relatively limited. This was particularly noticeable with the poor camera in an iPhone 3G – much less of an issue with better phone cameras. As an example, I could print a tag, photograph it (badly), display that sunken on a monitor and still scan it quickly and reliably with a the iPhone. Trying the same thing with a QR code failed even when much more care was taken to try to keep quality.
How is this important? Well, for marketers, an image that can still be scanned when it shows up at an oblique angle in a product placement or similar is worth real money.
I think the proprietry/open debate is important but not a deal breaker. History of technology is filled with examples of proprietary tech that becomes a defacto standard eventually without dire consequences.
The other point I think this article misses is the traffic monitoring and reporting that MS are providing for free which users would need to implement themselves otherwise with QR codes. Again, useful for marketing but also very useful for anyone else using codes to know if they are being used or not, and if so, by whom.
I think both options have their merits, and only time will tell which is the VHS/Betamax – but perhaps the most telling pointer is the lack of MS backing for their own tech, despite it actually being pretty darn smart (eg if tags were the open source option, I doubt there would be room for QR codes too).