I read it before (not sure if it was edited between the time I read it and the today), and my opinion is still the same and will not change anytime soon if the arguments also don't change:
Javascript sure is slower (and on what I saw, well, it is outdated and so weird that most people do not use directly Javascript but some libraries to help them). And it is an interpreted language which does not belong to a single interpretor, all of that is to take into deep consideration... unless you are not directly using javascript, in which case you should see what is already done to help.
In the case of C2, memory management and garbage collection issues are seriously lowered (due to general kind of recycling and layout by layout loading, as well as a lot of other things that I am not able to describe that may have been done), and the fact remains: Javascript performances are tied to how it is interpreted, on mobile I saw that, well, compatibility seems to be an issue so I would guess that we are far from what could have been done, yet, it is pretty doable.
All of this should be taken into consideration before buying C2 for obvious reasons (and you can test before buying it, C2 is fully html5 + javascript and does not rely on native at all, the event system should not be the only reason you are using C2 for, if html5 is not interesting to the potential user, then he might as well not buy it). If you need native, you won't even consider javascript anyway.
As for an html5 C3, keep in mind it is not before at least 1 or 2 years (and while JS performances does increase, as well as compatibility, which means it will be realisable, and before CC anyone though that a free drag n drop engine was just a toy, before C2 they though that html5 was doomed to failure and that nothing serious could be done with it, they might as well prove that a game creation engine can be done to run inside a browser without plugins) it is imaginable, heck even gdevelop as a web app nowadays (in a beta stage), sure it is not useable for a full fledged game and confusing, but this is because it is just weirdly designed, not because of perfs issues. However for obvious reasons when C3 does arrive, users should do the same they are currently doing: test carefully on a range of browsers so bugs are not left behind, also C2 already is a stable engine so the global design of it may be "easy" to transpose into the C3 editor.
however time will tell.
Ruskul I do agree that it would be nice to not have only javascript (due to how C2 plugins works, after all, there was a time when developping your own exporters was on the todo list if I recall), but well, no obligations, if other engines do better, fine for them.