I'm going out on a limb here and making a wild guess. "Academia" is too broad. Some fields may use one more often than another, but I really doubt that it goes much beyond a single author and how he/she is thinking at the moment. I think you will find wide variety of usage even in the same field.
However, I'll go out even further on the limb and suggest both a reason for this and what I consider a sensible way to write.
Sometimes you want to refer specifically to a particular paper and your phrasing or context makes that clear. Whether the paper has multiple authors or not isn't relevant. The paper is it and it is singular.
However, sometimes you want to refer more generally to the work of an author or a group of authors, of which a particular paper is only an instance. In such cases it is entirely natural to use singular or plural depending on the number of authors.
But it is a bit more complicated. You could, in fact, even refer to the work of a particular lab (in general) independent of its members, in which case, it is, again, singular (in the US, at least - see below), depending on your overall phrasing and context.
There is, I've noticed, a per-country convention in some of this. In the US, if I refer to the work of, for example, Google, I would use the singular (Google has produced...). But in the UK an organization is considered to be plural (Google have produced...). This same convention would naturally be applied to a lab considered as an organization.
So, my conclusion is, don't look for any consistency for Academia. But try to make clear in your writing whether you are referring to a particular paper (singular) or to the work, more generally, of its authors (varies).