It seems that some models of a multiverse actually are testable and falsifiable, and thus scientific. The notion of inflation leading to collisions with other universes may be one such model.
I have always disregarded the multiverse theory as false since it hasn't given valid falsification opportunity before. I think that the idea proposed is a step in the right direction and makes for the beginning of an interesting theory, but, obviously, it now must endure the rest of the scientific process. The truth will endure the trial unscathed, as always. Whether that means this theory will survive, we shall see.