Improvements to the indexing code could decrease that range of lambdas to about 200 supplied some hard work, so it seems to be like a hundred million documents employing the entire 1,000 lambdas appears to be achievable, but more most likely it would be near to 80 million when finished.