Stateless file systems perform interesting operations only when the mount point is crossed. Amd and all user-level file servers that I know of are contacted by the kernel only after the kernel has crossed their mount point during a lookup operation. Therefore, there is no logical place in a user-level automounter to call an operation when a mount point is crossed.
Since this type of file system is very limited in use and functionality, I will devote little or no effort to getting this case working in FiST.