I think I am on safe ground in saying that Dijkstra was well aware of the use of jumps in asm, but that is beside the point he was making in the body of the paper - as opposed to the point, taken from the title of the paper, that most of the argument has been about.