A good example of the HTTP client would be as a web crawler. This could also be a good demonstration of the flexibility of the URI class.
This sounds like a good idea. However I'm afraid of the effort that would need to go into something like this. Parsing HTML is scary stuff and finding the URL's and interpreting rel="nofollow" in a DOM along with the myriad other link elements in a document is a tall order.
That said, I'd be willing to accept contributions to this effect.
My motivation for opening this issue is to start generating more complex and interesting examples for the HTTP client. At this stage I am not worried about the complexity of the HTML parsing.