I am not in a field in which large (15-20+ authors) collaborations are common. However, I have always been interested in how authorship is distributed in those cases, particularly for extremely large collaborations (i.e., hundreds to thousands of collaborators).
How is it determined whether someone has had a signifficant contribution to yield authorship (i.e., why are there 436 authors and not 437 or 435)? How are authors listed (I know that LIGO, for instance, lists alphabetically)? How does one determine the major contributor? How is the paper drafted? Does everyone in the collaboration approve the paper?