At the PASS Conference in Denver the question was frequently asked - is it time yet for another release of SQL Server? Microsoft execs have said repeatedly at TechEd and at PASS that they've "heard us loud and clear" that 5 years is too long between releases. As a DBA with a company that maintains (and pays handsomely for) Software Assurance I understand that the 5 year window is too long.
On the other hand, I still have to support SQL Server 2000 because many of the applications we run are third-party applications, and the vendor(s) aren't yet ready to support SQL Server 2005. (I've even tested apps against SQL Server 2005 to show them it works fine, but they won't support it, so I can't upgrade.)
There are features in SQL Server 2008 that look like they'll save my company some significant time and effort. Things like Change Data Capture and the Dynamic Management Framework are really exciting. What I'm wondering is if it might be worth holding off a little bit to make certain that these work as they should work. It's hard to tell at this point, because of the slow release of CTP's for Katmai. (CTP5 was being called the "October CTP" at PASS. I even teased Paul Mestemaker about "October" meaning October 31. It's now November and CTP5 doesn't appear close to ready for public consumption.)
I'm well aware of the urgency at Microsoft on getting the code done and into the test cycle so the 2nd Quarter RTM schedule is maintained, but I also know that some features that should be in SQL Server 2008 won't be there because of that schedule. That's really too bad.
At this point I'm learning as much as I can about SQL Server 2008 because it's exciting, but it'll be at least two years before I'll be able to justify any upgrades here. The issue has more to do with my third-party apps than our in-house development, but it's hard enough to maintain two separate releases, much less three.