Improvements to the indexing code could cut down that number of lambdas to about two hundred given some effort, so it appears like 100 million paperwork utilizing the whole 1,000 lambdas appears to be feasible, but more very likely it would be close to 80 million when accomplished.