Crawler should fallback to the browser URL if the canonical fails

[leppert]

For instance, the canonical on this page throws a 404: http://www.scientificamerican.com/article.cfm?v=1&id=your-thoughts-can-release-abilities-beyond-normal-limits