That makes sense. But do your folks need to use persistent network drives? It would make things easier if drive mappings wouldn't survive a reboot.
Having said that, let's make a first attempt.
It's not easy to decide the subnet thing - how should that work? Should it be a decision based
- on IP address - dangerous if you use subnet masks differing from Classes A,B,C, or different subnet masks in the branches,
- or DNS names - difficult, as the full qualified name might not appear in net use, and a DNS query will not help.
The following snippet is based on
- each server is reachable via ping
- we are using Class C only
It does not matter if a drive mapping or only UNC is used. The net use /d is only echoed, for test reasons. You can remove the echo (word only) to execute it.