It's totally possible, they already have the infrastructure in place and 14PB of data available for download. Unfortunately the Wayback Machine data is not currently exposed publicly.
Then again, wouldn't it be pretty trivial to scrape?
(I say this as I'm working on a hellish scraping project, and the wayback machine seems like it would be a walk in the park to scrape)