It makes me a bit suspicious that a Canary build of Chrome is used. I understand that when a new feature is highlighted in a nightly build of Firefox, it's right to show it off - but why not do so against the latest version of Chrome - or even a Dev build? At least those two have been tested first ;)
Well, in this case it should be compared to whichever version of chrome performed best. That way you include any improvements they've made, but not any regressions.