Neither of the two parties mentioned is likely to resell the data directly, both are likely to to create a derivative work which they will exploit commercially

Neither of the two parties mentioned is likely to resell the data directly, both are likely to to create a derivative work which they will exploit commercially as will Google. What part of the ToS make one OK while and the other forbidden? How does this work when a smart scraper can just pull from the Google or archive.org cache?