While some of the results are interesting, I don't think this is a particularly good comparison. For a lot of the tests they said "This doesn't work on this browser, so we didn't include that test". Surely that should be a win for the browsers that DO support it, rather than just ignoring that feature. Personally, I'd care more that a browser can render more things, rather than if it can render some things a few seconds faster, but fail at others.
Not to mention, it completely ignores things like features, reliability, usability, security, etc, which are very varied between the different browsers. That's what I base my choice on anyway, and many that I've tried either crash, fail to load some pages, render pages incorrectly, or lack important features. Personally I find Firefox works best for me, but results would probably vary with different phones/OS versions, and some features are more important than others for different people
But hey, everyone loves benchmark numbers