The problem with this is that you end up with a beginner understanding of SQL.
For example:
* Your index then has only one field. GRATULATIONS.
* You can not use any of the advanced index functionality SQL Server has.
Not sure what you work on, but for me the whole generation is already unusable - I Run large data conversion scripts and generally use a lot of functionality that is clear and nice in SQL Server, but not exposed in any means. Among them filtered indices (common: 2 indices, same field, one non-unique for value null, one unique for value non null). What about included fields in indices?