The Schema Manager allows you to manage the update process (see below) and then perform the data structure (schema) update in your development environment and for your deployed applications.
The Schema Manager combined with the Metadata Manager (MM) holds all the information needed to create or update your application's data structure, in development and in a deployed app. The updater can even handle table and field name changes, and field type changes -- if you are coming from a previous version generated by the MM -- a special XML file is created in your database directory with the information from the last data structure update, providing the information needed to allow name changes.
The Schema Manager allows you to define update processes that munge your existing data before, in the middle (when the data has been moved over, but the new set of indexes hasn't yet been created -- important at times when creating new candidate keys) and after the data update, with the processing able to be geared toward certain versions (because, e.g., the data munging is dependent on certain fields being present, or not present, etc.). The Schema Updater uses those defined processes to smooth the way for your schema update.
In our VFP environment we also generated the SQL Script for MSSQL: this is plain vanilla, nothing fancy. If we had special requirements/needs, we would use the xCase MSSQL (or any other widely used relational database: it handles them all) to generate the script. We will bringing the MSSQL script generation, and data structure update process we use to update deployed apps running against MSSQL, over to this set of tools.