I am rather disappointed with the new pdf import filter. I imported a journal article from August 2010 and while the import itself worked, EN failed to extract all relevant information about the print version and only stated “2009” as the year (when the pre-pub version of the article was put on the web).
To document the case:
DOI: 10.1177/0894439309353037
Using a DOI resolver (dx.doi.org), this is the information available (screenshot : http://screencast.com/t/NzliZGU0ZT ),
but this is all what ended up in my EN record: http://screencast.com/t/NDJmMmI2
So, it would be easier to import from the publisher’s (here: Sage) database.