I am a fan of working papers. Berk Oezler has a post on the usefulness of working papers.
In my view, there are several advantages which can not be provided by reviewed paper journals.
1. Science is an iterative process and no result is perfect and final. This is the responsibility of the author
to present results as accurate as possible. Authors providing poor results become marginal quickly.
2. Double-blind review is a good thing to keep real findings far away from interested audience. I can refer to the complaints of the Nobel Prize laureat Prescott about the sad story of his publication with another Nobel Prize laureat Kyndland. The paper they were given the prize for was rejected by reviwers in many journals. It was first published as a working paper. Katarina Juselius also complains about enormous rate of rejections in American journals for european economists. Clear bias.
3. As a scientist, I am interested in new findings and ideas as soon as possible. I can judge myself what I need and what I do not need without reviewers, who might not be qualified to assess fine details. Working papers are usually published in electronic format which allows a prompt access and effective machine search.
4. Last but not the least. Articles in major journals are not free. I understand the reason. However, working papers contain same information for free.