I personally think like this:
If there is a new browser into the party, it should normally respect the same specifications as the others (since the others should respect said specifications too), so I do not think the difference of "working or not working" should be that huge. As for mobile or not mobile, why not, if it works and people like it, it is not like it is an issue at all.
As for the feature test, it indicates which features are supported by which browsers, not if any of them is fast, so a browser that support every feature / format / etc.. will not be faster because of that (it could still be).
On my part I saw that browsers could react pretty wildly depending on the system they run on:
-on my PCs, firefox is a pain, chrome is not as much a pain, opera was really fine but broken on some websites (according to other people using said computer).
-on my tablet and my phone, chrome is just not good enough (html5 games not made with C2 an be unplayable, and even when made with C2 it is a no go for my phone),, while firefox is fluid and works great for every html5 game I tried on my tablet (and was at least decent on said phone).
If we could have a browser that is not "the best for x, y and z platforms" but good in a consistent way across devices, as long as it is reasoneable, then maybe we could start thinking about it, right now, browsers are a mess (which is fine as long as you do not use them as wrapper, as the user choose the best for him).
Game design is all about decomposing the core of your game so it becomes simple instructions.