It's a least common denominator approach. If SSMS were coded to account for every permutation possible to arrive at the most efficient way to add a column think about the design, the testing and the support when new schema features are introduced. Now, all of that said, they have most of the technology worked out to do it and it has been part of Database Projects since VS2008 incl. VS2010 and SSDT for 2012, however afaik Database Projects still do not account for Replication. That is quite a tall task you're expecting. Maybe I missed it, but why even bother with SSMS if you know the issues and know how to work around it?