If you are not familiar with Evernote's Web Clipper, please install and have a look, it has a feature called "Simplified Article".
This basically removes all the additional components of the page and detects the main content of the page and shows in a simple layout. It then allows the simplified HTML to be saved on its own server.
I want a chrome extension that just does this. You might want to see how Evernote implements this by checking its sources and rebuilding that logic.