In addition to a more responsive interface, this dichotomy between server and client offers a lot of advantages. It means your backend is small and focused- defining a single entry point to your application’s data. The web client will interact with this interface but so can mobile and other third-party apps. Since the application and the application’s data are now completely separate entities they can now be distributed across the web as desired. The Unhosted project is doing some awesome work in this regard.
In my opinion, data-centric server backends like the ones described above are the sort of thing at which Node.js excels. It’s not uncommon for an application backend to be interacting with data in various places (i.e. databases, third-party APIs, filesystem, client connections, etc.) and this is Node’s strong point. Rather than interact with each resource one after the other (as is typical in most environments), Node.js can perform all these actions simultaneously and return the results asynchronously. This means both faster requests and the ability to handle more concurrent requests.