I think they mean the observable universe. I am no physicist and I only have a slight interest in quantum mechanics but I think the concept is roughly like this:
1. For every outcome in a closed system that is based on some chance quantum level event, the state of that outcome is fluid until observed.
2. It is the act of observing the outcome that places the human observer in one universe or the other.
In other words some event occurs on an infinitesimally small level that can cause two possible observable outcomes. Let's call them outcome A and B. In the writer's belief system the universe splits at this point into two identical observable universes except one contains outcome A and the other outcome B. The human observer is thrown by chance into either one universe or the other. Let's say he finds himself in the outcome A observable universe. At this point the human is completely cut off from contact with the other observable universe where presumably his identical twin lives with outcome B.
Wild stuff.
All this quantum stuff is fun to ridicule, but the weird part is that to the extent we’re able to test the hypotheses they turn out to be correct. A very great deal of our present technology, notably electronics, is based on it.
Sure quantum mechanics violates common sense, but this may be more an example of the limitations of common sense than of quantum mechanics. Of course, there are multiple competing and mutually contradictory versions of QM.
The more we learn about it, the more the universe, or possibly the multiverse, appears to be a very strange place.
I think this is a little different. I have never seen someone try to calculate the number of probable outcomes as a fixed value created at a fixed point in the past. I always read that the each observation by each observer in the set of all observers would fork a new branch. Therefore, it would have to be infinite.