I don't wish to hijack this thread, but I've been pondering a similar question. I've been working on a product that requires a very large amount of data that, as far as I can tell, can only be gathered by scraping (real estate data - even data vendors like estated.com don't have stuff like sales data).

Many, many websites contain legal language that forbids automatic data collection/scraping. How can a business be built in such a case?

Perhaps OPs tool only scrapes a select few sites that don't prohibit scraping, but that seems like the exception, not the norm.