It's not only about the client side. With the rise of node.js, it will be more and more relevant for server-side, data-heavy processing.

You are right, there are many things that can be done, using machine learning, like music classification by genre,artist or predictions api ( mentioned in a general fashion on a purpose because you are limited to the datasets and problems you choose to work on ). If a Javascript library that works smoothly and makes these things work better can be given birth i think it is definitely worth it. And you are right about the node.js thing.

I also point you out to some interesting resources about concepts that you could easily implement in your library. Like the NGD ( Normalized google distance ), just an idea, to make smarter tag clouds ? http://www.complearn.org/