This is a fascinating topic.
We have lumped together scraping in its entirety, even though the nature of the scrapings are quite varied.
I don’t believe Janet is saying it’s ethical, only that in the eyes of the law, taking a couple sentences from a site is not likely to hold up as copyright theft in court or in a DCMA complaint.
I have noticed that a lot of ‘scraper’ sites are scraping search engine results and are technically not taking content directly from other sites, but from the search engines. My concern with this is that the search engines don’t get confused between their own caches of site content and the original site content when handing out SE brownie points and penalties. However, when another site takes the same content and publishes it, the search engines do get confused and it may hand out a dup content pentalty to one of the sites, sometimes the original site.