In school I needed to scrape a few hundred thousand pages of a proteomics database website. For some reason you had to view each entry one at a time. There was IP throttling which banned you if you made requests too quickly. But slowing the script to 1 request per second would have taken days to scrape the site. So I paid <$5 for a list of 500 proxy servers and distributed it, completing the task in under half an hour.