I'm not exactly sure I understand what you mean... surely you're not saying that people's subjective feelings are more accurate? What would be a better way to test?
Have done things like set them side-by-side on the table and brought up browser windows and swipe screen to see how quickly the phone responds... That kind of thing.
And honestly the phones seem neck-and-neck. But this would be a good exercise to figure out how to better test this, because people just saying my phone is faster may not be the most accurate way to do it