Perhaps the word should be "notable" rather than "high-quality." A paper that is repeatedly refuted must have had some intellectual impact if others have bother

Perhaps the word should be "notable" rather than "high-quality." A paper that is repeatedly refuted must have had some intellectual impact if others have bothered to refute it in so many ways. PageRank would also take into account the quality or notableness of the refuting papers themselves, so this cannot be reduced to a mere count of the number of refutations. That means a highly-ranked, oft-refuted paper would have been refuted not just by anyone, but by other high-quality or at least highly notable authors.