New.FamilySearch.org (NFS) viewed each individual node as a cloud of individuals. All of the information about each node, accurate or not, was included in the individual's cloud. All submissions, no matter how far fetched were considered equally with all other submissions. There was no way either to focus the cloud of possible individuals onto the node or to eliminate any suggested individual from the cloud. So, rather than focusing on an accurate and source supported conclusion about differing information, NFS allowed the clouds to expand without any limits. So you would get individual submissions that were historically ridiculous that were essentially co-equal to the most source supported ones. Further, there was no mechanism in the program that allowed the users to focus or correct inaccurate information.
Family Tree eliminates those issues but allowing the information in the file to be edited by any user. This guarantees that the information in the trees will be more likely to approach the consensus of theoretical accuracy of any genealogical system. Although it appears that FamilySearch would like to focus on the cooperative nature of the program, in reality, it allows those who have the most correct information access to the focus point of the nodes by eliminating competing information. If there are real differences, based on different interpretations of the same information or conflicting information from different sources, then the Family Tree program provides a venue of the differences to be acknowledged and for the possibility of a negotiated truce.
Assuming that future developments of the Family Tree program as consistent with what has been represented as under development, then there will finally be a mechanism to "prune the tree" so to speak and eliminate nearly all of the extraneous cloud. For example, information coming from New.FamilySearch.org over to Family Tree might have a dozen or more name variations of the an individual. Here is an example:
These are variations in names for just one individual in Family Tree using the information that was automatically posted from NFS. The advantage of Family Tree is that all of those variations can be deleted from the actively viewed database. If there are actual variations, then the program allows for sources that support the alternate viewpoints.
This is a substantial step forward for FamilySearch. Many people, when hearing about Family Tree and the ability of users to change the information, immediately conclude that the information will be inaccurate. Family Tree's watch function will assure accuracy, if those with the supporting documentation will take the time to enter source citations and to communicate with submitters of potentially misleading or inaccurate information.
Since each person can only occupy one unique node, eventually, every person will be assigned a Family Tree node and the work of gathering further information will be concluded. Don't hold your breath.