I just talked with a man who interned with family search and who described, in a way I could understand, what's going on with Familysearch behind the scenes. Right now it's a bit like a baby in the womb. So that's the analogy I'll use.
The baby being the new feature "Family Tree" and the mother being "New Familysearch". If that means nothing to you I understand, new.familysearch.org has been present only in the backend and has only been available to premier members (kindof in a prolonged beta -- thus the "new" at the beginning). Family Tree on the other hand is a feature which at this moment is active and running and will eventually be available to the public.
(I think most would agree that new familysearch has kindof had a backend feel to it from the start)
Anyways! To push the analogy further these two features and sites are linked with syncing Java code. This means that at this moment if you change something in New - it will get changed in Tree and vice versa. The reasoning behind this is so that everyone can get a feel for how both work and make the transition seamlessly. But there will come a day when the umbilical cord will be cut and the code for Family Tree will have to stand on it's own. New Familysearch will become a thing of the past and this new feature will replace it. That is to come.
For the moment they are running synchronously. Tutorials on new.familysearch and Family Tree will be coming!
Check the post on Free Resources to find the links to get access to Family Tree. (or click here and follow the activation link for those with accounts)