> From what I read about the "theory of patches" I got the impression that
> it describes some (algebraic) group.
Not quite; multiplication in a group is total, while composition of
patches is partial.
Juliusz