you are going to want to scream about this question
this is an Oracle shop and will continue to be
all current environments are MSSQL 2000 Standard Edition
the answer(s) will help make a business case for upgrading
choice will be 64-bit and Enterprise Edition (leave that to me!)
end-of-life means nothing -- they have me (scary, isn't it?)
there is no development against SQL Server
there is (extensive) development with MS Access (recently 2007) that is only under duress moved to SQL Server
Windows Server 2003 operating system with no ounce of consideration of 2008
no data warehousing
no business intelligence
overwhelming percentage of accessing/supported applications are third party commercial off-the-shelf
scale of database size and usage negates the fact that you can 16-node cluster MSSQL 2008 instead of 8-node with MSSQL 2005
Here is the question, disregarding new features in MSSQL 2008 (which I believe to have a handle on), what significant improvements have been made in MSSQL 2008 to features that exist in MSSQL 2005 and not in MSSQL 2000?
Page Compression. OK, so is a new feature, not an improvement of a SQL 2005 feature. You can spin it as is an improvement over row compression which exists in SQL 2005 SP2. Besides, it affects, transparently, almost every feature in SQL 2005, no changes required. Its such a benefic feature that it justifies SQL 2008 over 2005 from every possible angle.
Significant is a highly subjective word, but off the top of my head, one of my favorite improvements is the separation of Date and Time into separate datatypes instead of one datetime datatype. Also, the range for dates now goes back to 1/1/0001 instead of 1/1/1753.
I'm sure there are performance improvements, etc, but from my experience, 2005 was so solid that upgrading to 2008 is more about the new features than the improvements. In our shop, we are upgrading only when we can take advantage of the new features (e.g. filestream).
I think the question that you're asking isn't really the question. I work for state government and have worked in the private sector. Government usually has longer cycles for projects, but the core problems are the same. If you work in an Oracle shop, pushing a Microsoft, IBM or MySQL solution isn't going to be productive or good for your career. Unless you have solid backing from reliable management types, you're going to be perceived as "one of them" and will have a hard time getting things done.
In your situation, I'd push to treat off the shelf apps with MS SQL requirements like appliances, with small-scale databases associated with the app. You may have more success getting what you need in that case, as the Oracle guys think that MS SQL is a toy anyway.