In doing some refresher reading on Everett today I am reminded of a very important point. Everett has many supporters among quantum mechanics physicists. But that support is split into two camps. Most recognize that many worlds is a useful language for simplifying quantum mechanics but don't view these other universes as "real" in the same sense ours is. They are paper universes (kind of like square root of negative two is mathematically useful but you won't run into it as a physical object in the real world). Stephen Hawkings is an example of this camp.
The minority of many worlds supporters hold forth the idea that these other universes are just as real as ours. DeWitt and Deutsch are in this camp.
And of course there are many equally prominent quantum mechanics physicists who consider many worlds to be complete hogwash (polls tend to find about 50% accept some form of many worlds theory). As
this Martin Gardner column points out, Roger Penrose is in this camp.
Everett's many world theory, where infinity allows all possible outcomes to be realized in proportion to their probability means free will can not exist (since if all outcomes occur then no "will" was exerted). I know some proponents of Everett's many worlds disagree (such as DeWitt; but there is almost universal agreement otherwise that a deterministic multiverse destroys free will), but I'm of the opinion that they do so by changing the definition of free will away from what most people mean by it.