Very complete article, but there seems to be a lot of re-inventing the wheel here. We already have GIS (Geographic Information System) software that does automatically (in the background) much of what had to be coded. The article also doesn't mention using GIS data already stored in SQL Server, i.e. in an SDE (Spatial Database Engine) geodatabase. While this is specific to ESRI, a very large chunk of GIS data in many organizations is stored that way.
Another concern is turning database people untrained in cartographic principles loose on mapmaking. I have seen extremely misleading maps made by very well-intentioned folks. A good book about this issue is "How to Lie with Maps" by Matk Monmonier. It is very easy for those untrained in spatial thinking to apply the wrong processes to solve a problem, resulting in totally wrong conclusions.