There is nothing efficient about storing a document's data stream in a RDBMS table. The name we all give these database-bound streams of data is BLOB, for Binary Large Object. They aren't relational in any way, you can't filter or sort on them, so why are they there?

BLOB I/O consumes a considerable amount of a SQL Server's CPU and Memory resources, leaving the SQL Engine with less resources to perform queries and transactional data operations.

Even if you could overcome the BLOB I/O challenges by investing more in your SQL tier, you're still left with the need to make considerable investments in expensive storage to address performance and/or disasater recoverability concerns.

BLOB Remoting removes SQL Server from the BLOB I/O workflow and in turn rids you of BLOBs altogether, as depicted in the following diagram:

If SQL Server is not involved in managing the I/O of documents stored in SharePoint then a BLOB is never created...SharePoint is BLOB-Free! You're left with an implementation that will perform better, more easily and economically scale, and be simpler to administer and maintain.