PDF seem to not rank as well as other normal webpages.  They still rank do not get me wrong, we have over 100 pdf pages that get traffic for us. The main version is really up to you, what do you want to show in the search results.  I think it would be easier to rank for a normal webpage though.  If you are doing a rel="canonical"  it will pass most of the link juice, not all but most.

PDF seem to not rank as well as other normal webpages.  They still rank do not get me wrong, we have over 100 pdf pages that get traffic for us. The main version is really up to you, what do you want to show in the search results.  I think it would be easier to rank for a normal webpage though.  If you are doing a rel="canonical"  it will pass most of the link juice, not all but most.

Indicate the canonical version of a URL by responding with the Link rel="canonical" HTTP header. Addingrel="canonical" to the head section of a page is useful for HTML content, but it can't be used for PDFs and other file types indexed by Google Web Search. In these cases you can indicate a canonical URL by responding with the Link rel="canonical" HTTP header, like this (note that to use this option, you'll need to be able to configure your server):

Link: <http: www.example.com="" downloads="" white-paper.pdf="">; rel="canonical"</http:> 

Google currently supports these link header elements for Web Search only.

You can read more her http://support.google.com/webmasters/bin/answer.py?hl=en&answer=139394

How did it turn out?  And Mark have you done much with internal search?