the fun of a key-value datastore! i don't know of a way to copy the data with web2py "magic", but you can write a migration script that does that for you, and then as you create new tasks you can write all the data necessary at task creation time.
another thought - i doubt users can use over 1000 records of either at one time. try using limitby in your queries to limit to data that will be consumed. you might consider catching (partial) result sets from the less-dynamic of the 2 tables. if you are dealing with background processes, look into "backends" that have unlimited time to process a request.
