The endpoint of the Gregory-Laflamme instability of black strings revisited