Here’s an interesting piece from George Lozano showing that, while journal impact factor became an increasingly better predictor of citation rate of a paper until 1990, over the last two decades, that relationship has begun declining. Figure 1 (below) is particularly interesting.