Why would anyone believe these results? Without anyone looking over Aruba's shoulders, they could have done any number of things to engineer the results. At least people like Tolly make sure there is no dishonesty and ensure reproducibility.

Come on, no one gets 3 Mbps. They clearly are doing something to the setup to produce those results, and I am not surprised that they maneuvered to put Meru as the worst, because they are clearly the most afraid of them. Did you see their other whitepaper trying to trash Meru's architecture? The effort is clearly unscientific, like a drug company making a new and expensive painkiller trying to show that aspirin is fatal by comparing aspirin with the infirmed to the new painkiller with the healthy. John Dunn has a follow up article with Meru refuting Aruba, but it is not on Network World yet.

And the Cisco results are just as ridiculous. Cisco is a part of the WiFi tests, and Aruba isn't, so there's no reason to believe that Cisco doesn't actually work better than Aruba. And Cisco has its CCX alliance, as well as a partnership with Intel.

There are client differences now but they will go away with better software tuning on the client and more appropriate TCP settings. It doesn't help anyone for Aruba to try to trump up their own tests.