Have you tried this in CLI? I think the problem lies in that you've give only the symlink user access rights. For instance, let's say that A is the actual destination point of symlink B. So:
B -> A
Now, currently permissions are set as follows for the files (example only):
A - rwx for owner, r-- for group, r-- for world
B - rwx for owner, rwx for group, rwx for world
So, anyone can read, write or execute file A but only the owner (root in this case) has full access to it and everyone else can only read it. Since file B points to file A, it's just redirecting all users to A. However, no user but the owner (again, root here) has permissions to do anything but read it--but everyone has permission to be redirected by B to A. It's like everyone gets slingshot from the pointer B into a brickwall A. So, what you should do is set the permissions for the file A so that the users can do more than just read. However, it depends on just how much permission you want granted. You could give everyone read, write, and execute access but that could pose a security risk.
In the CLI, su - to root, then set the permissions as follows:
chmod 755 <filename>
but without the <> and obviously replace "filename" with the actual filename. This will give the following permissions:
owner (root in this case): rwx
If you need more information, man chmod will help you out.