The point of this post is to introduce the term nonspecific citation.
A common problem with antibody is non-specific binding: while the antibody may be targetted to grab on to protein P, it actually also grabs X, Y, and Z, which are somewhat similar to P.
A nonspecific citation is when paper gets cited as an example of a broad class rather than for its specific ideas.
In the case of a review article, it might make some sense: if you are just mentioning the field, you might as well cite my review of bioimage informatics. Even here, though, you could cite other reviews.
In the case of the research articles, though, it is often more of a throwaway citation: we need to mention work in the area of cell image analysis and I have a fairly recent paper on this, so you cite that. Or you cite somebody else’s work as it does not really matter the specific contents of that citation.
There are intermediate levels of specificity.
Maybe you are writing a paper that somehow touts the general usefulness of local features and you write a sentence such as:
Local features are useful (or SURF has been used) in many context such as A [paper], B [paper], and cell image analysis [my work].
Then there is the completely nonspecific:
Recent work in bioimage informatics […, my work, …] …
This will of course lead to a citation Matthew Effect: if you are highly cited, you are likely to get even more citations by getting more nonspecific citations.
Whenever I have a paper I really want to read that is not open-access and to which my institution does not have access, I generally get it by asking the author for a preprint. It has never been a problem and has even led to some good follow up discussions.
On the other hand, if I am not looking for a specific paper but just something in a general area to get acquainted, then it whether something it open access or not may make me choose to read (and then subsequently cite) one paper instead of the other. In fact, just seeing an IEEE link will be enough for me to not even click through to the abstract.
Maybe if open access does have an advantage in getting citations, it’ll mostly come in the form of nonspecific ones.