Indeed - and i believe scraping data off a publically accessible webpage (e.g., one where you do not require registration and login/password) falls under fair use, provided you do not take up more bandwidth/resources than the average user of that site.

I recall there being some precedent for this sort of fair use - something like a phone directory - the information is not copyrighted, but the arrangement and layout is. So hence, you can't just iframe a site and present it, but obtaining the data, and deriving a new work from it should fall under fair use.

That court case was Feist v. Rural. The court ruled that copying the listings in a phone book was permitted because facts cannot be copyrighted.