It may technically fit into some people's definition of a "scraper site", in that it does not provide any human input, and it is automated. However, it's provides an extremely useful service to me. The algo they use is great. I love being able to read 5 different takes on the same story (or 500 if you wanted). I've never found a scraper site that's remotely useful. Plus, G hasn't really monetized Google news (as of now). Scraper sites always are as spammy as can be. Interesting question you bring up though.... Trying to filter out scraper sites will definitely be tricky for SEs. News is so syndicated these days, not much is original.
Topix.net is as scraper as they come and it was just bought for $50 million by a group of various newspaper companies. I'm sure they are making quite a bit through Adsense.