Darragh,
Remember that Windows 2000 is based on the NT kernel. I won't say the two tips I mentioned in the article are the only ones for this problem, however, they've worked in most of the situations I've seen.
As far as the shoulder shrug goes, well, that's the nature of the development process. After all, what's more important? Pulling an old build out of the archive and updating it to work or making sure the problem is fixed in the present version? I can certainly see two sides of the problem, but given that Paradox 7.32 was relelased some five years before Windows 2000, it's pretty easy to make the case that there are diminishing returns--especially when you're development staff is limited and there are far larger issues in the current code base (Example: Printing).
Also, given that there are workarounds available, it does make a certain amount of sense to lower the priority of that one--especially given the twisted nature of the Paradox 7.32 patching process (which, to my mind, was the most difficult patch to apply of the bunch).
Yet another thought to think about...
-- Lance