Jon@tchambers as the linked article says, companies *don't* base their claims on ownership; in fact, they "expressly disclaim any property rights in that data in their terms of use."
But, I agree that scraping hasn't been tested for anything federated. And in a situation like indieweb.social where the instance admin is publicly saying he doesn't think there any legal barriers to scraping data, courts might well find that companies can rely on that.
All GNU social JP content and data are available under the