We have a large digital library with thousands of papers indexed in Google Scholar. We allow Google Scholar to index our PDFs but they're blocked unless you have a subscription. So Google has full-text indexing/searching of our PDFs (great!) but then the links point just to those PDFs (boo!) instead of the more helpful abstract pages.

Does anyone know what could cause an issue like this? I am, to the best of my knowledge, following all of the guidelines laid out in their Inclusion Guidelines.

Here's some example meta data:

<meta name="citation_title" content="Sample Title"/>
<meta name="citation_author" content="LastName, FirstName"/>
<meta name="citation_publication_date" content="2012/06/26"/>
<meta name="citation_volume" content="1"/>
<meta name="citation_issue" content="1"/>
<meta name="citation_firstpage" content="10"/>
<meta name="citation_lastpage" content="20"/>
<meta name="citation_conference_title" content="Name of the Conference"/>
<meta name="citation_isbn" content="1-234567-89-X"/>
<meta name="citation_pdf_url" content="http://www.example.org/p/1234/proceeding_1234.pdf"/>
<meta name="citation_fulltext_html_url" content="http://www.example.org/f/1234/"/>
<meta name="citation_abstract_html_url" content="http://www.example.org/p/1234/"/>
<link rel="canonical" href="http://www.example.org/p/1234/" />

example.org/p/1234 is the abstract page for the article; example.org/f/1234 is the fulltext link accessible to subscribers only (and to Google Scholar). example.org/p/1234/proceeding_1234.pdf is the fulltext PDF link.