The question in the title of this post is easily the most repeated question I hear in class after class, in fact I was asked the question yet again, just a few hours ago. I have a long explanation for why FamilySearch.org's Family Tree program does not "work" but the question addresses multiple layers of issues and the answer is quite complicated. But before getting to the heart of the answer to the question, if there is an answer, I am going to address several issues that are not symptoms of the underlying problems.
1. The Family Tree program is not broken and does not need to be fixed.
There are two parts to the issues surrounding Family Tree; the program itself and the data that went into the program. The program is the structure. The data is the content; consisting of over 150 years of accumulated, user submitted, unreviewed individual and family information mostly without sources. The program works very well and is the solution, not the problem. The data is the problem (or challenge, depending on your point of view). The program continues to evolve as features are added and changes are made. This is a natural consequence of technology and computers. Blaming the "program" or blaming "FamilySearch" for the failings in the data is short-sighted and unproductive.
2. Changes made to the entries in the Family Tree do not indicate that the program does not work.
To the contrary, the fact that changes can be made by anyone using the program is the main indication that the program works as it is supposed to work. The program is designed to be entirely collaborative. It is a wiki and wikis work to handle huge amounts of data such as the entries inherited from all those contributors for the past 150 years. FamilySearch is merely the messenger, not the message. All of those seemingly random or inaccurate changes to "your data" are really the combined contributions of all your own relatives. To quote President Harry S. Truman, "If you can't stand the heat, get out of the kitchen." Or in other words, if you can't stand to help correct the data from your own relatives, don't complain about the task. Leave it to those of us who understand the task and will do our job to clean up the data (I was going to write "mess" but thought better of it).
3. FamilySearch is not making changes to the data in the program.
I have to admit that FamilySearch leaves little calling cards all over the data, but ultimately, as I explained above, the data comes from our own ancestor's contributions. That fact that they are delivered with a calling card emblazoned with the name of FamilySearch is just an unfortunate artifact of the way the data is being parceled out into the program. The original data comes from several large accumulations: the Ancestral File, the Pedigree Resource File, the International Genealogical Index, LDS Church membership records, etc. Additional information has been added by users over the past years including random GEDCOM files. All of this was used to create New.FamilySearch.org and then passed through to Family Tree.
4. Truth will prevail.
Genealogists are not used to working in an environment where other researchers can have instant access to their data. They usually work in a vacuum. Family Tree is the exact opposite of a vacuum and is, for many genealogists, a new and very uncomfortable environment. There is really no other way to rectify the past genealogical confusion. This is really the first time that there has been a glimpse of a solution to counter the actions of each genealogist, good, bad or indifferent, from doing exactly what he or she wanted to do without even the slightest opposition or challenge. But in the end the persistence of sources will carry the day. Family Tree is source-centric. This means that entries must be supported by sources to survive the scrutiny of other researchers. Unfortunately, there are a huge number of users of the program who are learning this fact the hard way; they are having their entries changed for no apparent reason. The program prompts users to make changes only with an express reason. Presently, changes can be made by anonymous users. This is a program flaw that needs to be corrected. No one should be allowed to make changes to the program without providing a way to allow others to communicate with them. There should at least be a way to pass messages to all of the users, even if they do not have contact information showing in the program. If the program allows a special category of users, ones who can make changes without accountability, then there will always be random and very bothersome changes being made. This issue has been discussed repeatedly over the past couple of years and still is not resolved. This is the second most common complaint I hear every time the subject comes up.
5. Sources are the solution to nearly all of the data problems.
In the end, when all is said and done, it will be the quiet, persistent addition of sources that will quell the rabble and give prominence to careful, reasoned, historically correct data. Rather than complain, save the energy and put it to use adding even more sources. There are still vast areas of the data where no sources exist. None of this data lacking sources can be verified or relied upon until the work is done to add sources. Cleaning up the data and then maintaining it in response to new sources is the ongoing work of the Family Tree.
The question still remains as to when all this will come to an end? When will the problems created by the data end? To some extent, as long as there are careless or ignorant contributors, the problems will always persist. But in a real sense the problems caused by this element will diminish dramatically as the data is finally completely converted from New.FamilySearch.org and any other issues remaining such as those involving membership records etc. are resolved. How long with this take? It will take as long as it takes. That is the only possible answer.