My stance on native engines is detailed here: https://www.scirra.com/blog/ashley/28/the-case-against-native-engines
I think "make native engines" is everyone's knee-jerk reaction when something doesn't work. For example, sure, we could make changes to make publishing easier. But that can be done anyway. You don't need native code to solve that. I also have direct experience of working with a native engine for Construct Classic. That had enough problems to make me move away from it. It is certainly not perfect. Do not imagine it will magically fix everything!
I have always believed that javascript is very slow and i believed that Native is the best solution but game design is also a huge factor (Actually the highest factor for me).
But I imagine in a perfect world considering that game design is in it's optimum condition that going Native is still better in theory. Until your blog post. You have mentioned about WebAssembly, by the name it's obvious to me that it is also a low-level programming language like Assembly and generally way faster and might even clear my doubts.
But also as you said, it is still in development.... Not to mention, it is also unstable. According to Google anyway...
My question is, would it still be helpful to C3 on it's current state? Waiting for it may take too long... or am i missing something?