The question you should be asking is what is so different about your PDF's that they should be treated differently than other PDF's? Since a PDF can be anything, yours are not special.
What is breaking the user experience of browsing the web these days, is that many websites treat links differently. There is nothing for the user to learn, there is no mental model to be built, from the user's perspective what happens when you click on any link is completely random. One reason for this is that people let their personal preference get in the way of designing a proper interface. I see lots of people that prefer PDF's to be downloaded and make that the rule for other people regardless of what they think. I've had many clients request that links to anything but their website is opened in a new window, but if you fear losing visitors taking them hostage is not the answer.
The only way we can make people experience a sense of autonomy and control when browsing the web is to stop customizing what clicking a link does. For your visitor, your website is just like any other website. Please just let them use, and learn to use, the control that is already at their disposal. Right click in any browser allows you to download a PDF instead of opening it (that is, unless you broke the behavior of the link). Any PDF viewer in the browser has a quick "download me" button. While Nielsen and Norman are probably right, it should be interpreted as aimed at builders of browsers and PDF readers, not at individual website developers.
A bit of personal perspective. If you're doing a literature study that involves looking through dozens and dozens of research papers, the last thing you want to have to do is to keep track of your browser, your downloads, and your PDF viewer. If you're unsure the PDF has what you're looking for it's much much easier to view the PDF in the browser, have a peek, hit the back button. If you need to recall it it's in your browser history, if you don't it's gone. If you downloaded it, you'd need to wait for the download, find it in either the download manager or Windows Explorer, then open it in a reader. If you find you don't need it, you need to close it, delete it, find your way back to your browser.
Literature research made me love the browser-PDF integration. Entire populations depend on it, don't break it.