Is there any fundamental difference between mapping a network share and using a UNC shortcut in Windows (aside from the obvious loss of a drive letter)?
In practice, there are fundamental differences between mapping and using the UNC path.
Some programs (mostly old) can't work with UNC shortcuts. There are also scenarios where the use of the UNC path is recommended, e.g. programs that run as a service. So you've decide in a special situation what's the better or more compliant solution.
In theory it should be the same, but in practice you may encounter applications that assume there is always a drive letter. These sort of apps make it difficult to use a UNC. Older apps written with once popular toolkits (e.g. Borland's widget set) are a good example.
Without access or knowledge to the Windows source code, this is difficult to answer.
Network drive: Compatible with offline files (Windows feature) Can cause slower logins Compatiple with leagacy programs Easier for users to understand (IMO) Useful for distributed file systesms that cannot hide server details very well Easier to index / use with Windows Libraries Shortcuts Initial access can take long (kerberos auth / name resolution) Personal favorite
Don't believe the "it's easier to change server details with mapped drives via group policy" you read on the net. Last time I checked GPO could control shortcuts as well. My rule of thumb is to use shrotcuts whenever possible.