I say change the IP schema on the PIX side. If you are on DHCP, that makes it very easy for the clients. Then all you have to worry about is the no-brainer VPN. Hope his helps.
Changing the schema is easily said, but difficult to achieve in a lot of circumstances. A much easier way, and all under your control, is to just NAT the IP address of the server, and of the source network.
Then the server just thinks it is talking to a different subnet, and the users access the remote server on a new IP address. The VPN access list is then configured to use these natted addresses rather than the ‘real’ ones.
I do this all the time, looking after a VPN netowrk with about 30 remote sites (Customers), and many of them have the same IP address schema. Using NAT hides all of this from the users, they just think there are many different subnets, and the ASA does the rest (with a little help from my config !).