MDPI is gamed by design, I think that while Elsevier is awful, MDPI is even worse with 100s of special issues where you are guaranteed to land publication in journals with quite nice IF (which is inflated by publishing large proportion of reviews and less original research).