The main idea of this paper is that the old ways of determining credibility, by checking sources, is becoming obsolete with the rise of the Internet and particularly collaborative sites such as Wikipedia. Instead, users find various cues on the site which give an overall impression of credibility. For example, on Wikipedia there is recent information, many images to complement articles, and external links, all of which make the site seem more credible. The perceived credibility of a site like Wikipedia also depends on the users' view of the process by which its content is created. A user who does not believe that a collaborative encyclopedia is feasible is not likely to give much credibility to the information in Wikipedia.
I think that source credibility actually can become an important factor though. A distinguished website or the author of a popular blog can hold credibility on their own. People may think an article on Wikipedia is credible simply because it is on Wikipedia. Basing the credibility of some information on its source is a way to avoid the work of analyzing it. I think it makes as much sense to attach source credibility to a website like Wikipedia as to a person. Both are, in a sense, distributed knowledge networks. One is made of webpages, while the other is made of neurons. Actually the distinction is fuzzier than that, as people rely on the external memory of papers and websites, while human contributors are integral to the knowledge network of Wikipedia.