I am reading R. Austin Freeman's Eye of Osiris which so far is a pretty decent mystery. The quality of the writing is not bad, though not so great. It is unduly prolix at some times (I suppose one could say that that makes for some character development of the narrator, though, but that could still be done otherwise), and occasionally at the beginning there is a lack of clarity. This brought home to me the obvious fact that the skill of story-making is very different from the skill of story-telling. While there may be some correlation between the two skills, I do not know that the correlation is very strong. In any case, one would expect that there are a number of people who would be excellent at story-making but whose story-telling is subpar, and a number of people who would be great at story-telling but only if someone else made up the plot. But now the puzzling fact is that there are very few co-authored novels, and off-hand I can't think of any famous ones written by two authors working together (a number of works do draw on older texts or traditional elements, and maybe that should count co-authorship, but I don't want to count that as "working together").
There are artistic genres where collaboration is routine. Film, theater and music perforances are obvious cases. I do not know if Greek statues were carved by one person and painted by another, but it certainly would make sense to have that sort of division of labor. On the other hand, my sense is that painters, writers and contemporary sculptors tend to work alone.
Here is a hypothesis: The artistic vision tends to be hard or impossible to communicate except by means of the genre of art in which it is to be embodied. But collaboration would require communication of the artistic vision. And that artistic vision cannot be communicated except by the work being produced, which presents a vicious circularity in the case of collaborative works.
This argument can't apply always, because there are collaborative works. But maybe we can say something about these exceptional cases. I don't know if the painting and the sculpting were separate in Greece. But if they were, we might say that the sculpting was to some degree an independent work of art, which communicated its vision by itself. After all, most of the Greek statues we see in museums have lost their paint, and yet the sculptures appear to us to be complete works of art—so much so, that in times past people didn't seem to know they were originally polychrome.
The case of directors and conductors is a bit different. But there we can take the director or conductor as in some important sense the author of the performance as a whole. There is a communication between the director and conductor and collaborators, but in an important sense this communication is itself essentially a part of the genre (cases of films written, directed and starred in by one person are defective cases of the cinematographic art), in a way in which communication between authors is accidental to the novel. Moreover, note that the director and conductor only communicates to each individual collaborator a part of the artistic vision.
So this might explain why a superlative novel is unlikely to be produced by two or more authors. But still, a decent novel could be, and sometimes is.