If one wants to separate the question of scraping from the question of publishing something scraped (Which are entirely different things) then the thought experiment should be e.g. that you scrape a catalog and your page only displays "right now there are 10224 articles in the IKEA catalog" or "right now 1234 bikes are for sale on craigslist".

That material has been reduced to a number - which certainly can't be considered protected as a material.