Multiples of Ambiguous Positions
Bill: A comment by Robert Pauli led me to take a look again at multiples of ambiguous positions, and to realize that I had made a mistake on Method of Multiples, when I said that you could take the fact that there is never a miai, no matter how many copies we have, as a definition of sente. That is also true of ambiguous positions.
The first kind is like a sente, except for privilege. Since the threat (follower) is no larger than the play itself, the play cannot be made at a higher temperature, when a reply is bigger than anything else. So the odds that the opponent can play the reverse sente approach 50-50. Still, it is like a sente in that, if the play is best, the reply, being of the same size, is likely to be good, too. Therefore, some people have regarded such plays as sente. For them, the method of multiples justifies that.
What I had overlooked was the second kind of ambiguous position, which also does not produce miai with multiples. Suppose that there are several of them. If the player with both options (player A) plays first, she will play all but one with sente and then play the last one with gote. If the other player (player B) plays first, she will play it with gote, but then player A will play all but one of the rest with sente and then end with gote. The result will not be miai, OC. In fact, it will look like the result when player B has sente!
So it seems like only true gote produces miai with multiples.