Let's try to imagine how this would look if we changed the almost everyone to just everyone.
Last year, everyone in the editorialist's country who claimed in polls to be tired of television news programs' obsession with celebrity scandals and to be unwilling to watch such programs continued to watch those programs.
I still dont think this is a necessary assumption. We do not know who was watching the shows, or how many people watched them. lets say that 60 percent of the country were tired according to the poll- but the soaring ratings was due to a rise from 10 percent to 20 percent of the country watching. In this case, it is possible that we are still talking about two completely separate groups of people.