I would go with the screen scraping method like hex1848 posted.. it can get problematic if the links and images have relative paths instead of absolute... i've done it but its a big pain in the ass to translate the relative paths...
http://www.thegopedindex.com/demo
Its a framed page but just for looks, notice all the links and image paths, everything is being pulled from the site by a com object and then parsed through before it is spit back to the browser.