I know this is probably hugely unpopular here, but I really wish coffeescript would go away, but perhaps some of it's best features get integrated into the next version if JavaScript.
I use Backbone.js quite a lot, though, so I suppose you could say I am a fan of Jeremy Ashkenas' work, even though I don't like coffeescript. It doesn't solve anything for me - it just makes things a little different without offering any major benefit for me. The article you linked to was an interesting read, I hadn't really thought about it enough to put it into words.
I don't really like being a naysayer about technology or sticking my head in the sand. I think the best parts of coffeescript will eventually make their way into Javascript and then coffeescript won't be necessary anymore (unless it also continues to evolve). In which case, browser support for it seems unnecessary to me.
Or, perhaps it'll become a first class language in the browser and then in two years we'll have some other language that compiles to coffeescript...? It seems ridiculous to me, but I don't see why that wouldn't happen.
> I think the best parts of coffeescript will eventually make
> their way into Javascript and then coffeescript won't be
> necessary anymore [...]
If javascript was good there would be no need for all these langages that compiles down to javascript.
Anyway, i did not not the "then" trick with classes , like
new class then constructor:-> @param="something
Coffeescript feels to me like the icing on the cake for anyone who understands what the resulting JS will function like.