A patent examiner finding an academic article as prior art that is relevant to a patent application is faced with a strange situation (at least in Israel, where I work). The article will usually cost around 30$. Either the patent office will pay for it (using tax money, something examiners are not allowed/encouraged) or the applicant will pay for it once the examiner cites it (since he doesn't really have an option not to). Many times the citation consists only of the abstract (since it is visible for free). Another common solution for the examiner is to just not cite the article (since the examiner knows that citing only an abstract is a weak citation), and then we all loose: the applicant gets a weak patent, and we all pay more for another monopolised invention that is not new/inventive.
The problem raised is not unrelated to a different comment
Rather than the 8.5B or 2.5B cost why not just publish your dam [blank] with a free online software with collaborators that you can invite in to edit?view?private?public?ect ect ect.
But the patent examiner cannot utilize random "publications" found on the internet. Not even supposed to use Wikipedia. So the lousiest formally published journal carries more weight than anything on the internet [?] Further, with the changes in Google in the last month or so, it's harder to find content.
[The Google ranking system, as to IPBiz, changed considerably between Feb. 6 and Feb. 13, and more so by Feb. 20.]