You're responding to a power efficiency result by citing a pair of raw performance results. That doesn't make a whole lot of sense, especially when discussing mobile hardware. Differences in power efficiency usually matter a lot more to mobile hardware than differences in raw performance.