Leigh,
No -- I haven't progressed on it further and don't expect to.
With any kind of automated translation project like this you hit the point of diminishing returns, and after I taught myself the basics of parse tree rewriting and got the acs-core done, I hit that point. Anything beyond that would just have created a maintenance job (keeping pg+mssql+oracle in sync) that neither I nor anyone else really wanted.
If I were looking to port a great deal of the acs kernel code (or beyond) in the future, I might look to some of the vendor tools for doing this. Microsoft now has a free Oracle Migration toolkit, for example, (http://www.microsoft.com/sql/solutions/ssm/ssmav2.mspx) and other commercial ones exist. You might try begging a license from them for use in open source projects.
John
