Searching using the index.pdx will only function if the user can map to the drive the index is stored on. If you plan to use this via the internet you are really out of luck. There are tools out there that will allow you to build pdf search indexes for the web if cost is not a factor. Hopefully, the next version of acrobat will address this issue.
Now, if you plan to allow users to search the pdf meta data only (author, title, subject and keywords) you could extract this information to a database that can be called based on your scripting. We are in the process of doing this since our users use these fields for all their drawing retrivals.
To extract the meta data you could write some code to this. We use PDFWorkshop to do this since we use this tool to manange the metadata.