I realise this is has been around for a while now but until a few weeks ago I never really appreciated the Group Policy Preferences and the simplicity they offer.
Back in the days of Windows NT, Server 2000 and Server 2003 server administrators would create login scripts to perform a number of commands such as mapping network drives, installing printers, creating shortcuts and folders… I could go on but you get the idea. In Server 2008 Microsoft introduced the Group Policy Preferences to allow you to use Group Policy to natively configure a whole host of setting in Group Policy that would otherwise be a number of lines of batch/kix/vb script.
As you can see from the image to the left there are a vast number of options that can be configured for a user when they login. For most of the items there are four options: Create, Delete, Update and Replace which will let you make changes to the Drive Mappings, Folders etc. The difference between Update and Replace can vary from item to item but my general understanding is that the Update will attempt to modify the existing item to match what is in the Preference whereas the Replace option will remove what is there and recreate the new object (smilar to a net use P: /DELETE /Y followed by net use P: \ServerUsers%USERNAME%)
Another benefit is that in a single GPO you can define a number of different Preferences and then filter these around Group Membership.
This should all work Out of the Box with Windows Vista and above so for any legacy clients and servers (Windows XP, Server 2003) you will need to download the appropriate updates from Microsoft http://support.microsoft.com/kb/943729.
All in all this should save time and administrative overhead when they are fully adopted. The only problem is getting the legacy scripts switched over to the new Preferences.