the document is not an essay explaining how the fear of an overwhelming number of incompatible forks of a project is not upheld by reality.
It seems to me like that's clearly what it's about: it gives several examples of instances of forking in many other projects that have been perfectly fine, and in the final section explains why that has happened. In the BSD section, it explicitly talks about how yes, sometimes forks persist, but it's because they fill a specific niche, creating a situation where it's clear which fork to select based upon actually substantive differences.
it is about the legal fear of forks. as in incompatible licensing and cetera. there's a reason it's under a subdirectory literally titled « Licensing and Law ».
Eh, death of the author. I don't think that what it says on the legal issues of forking are the main thing worth taking away from it.
the document does not address the psychological overwhelming of many forks to choose from, which is only a counterargument addressing your argument that many modifications to choose from is overwhelming.
I think it addresses that by explaining that forks rarely become many.