Averages that include all US patent litigation (such as the ones you cited) would seem to be unreliable for a case like this one that has so much money at stake. Many patent cases have minimal commercial relevance, which intuitively ought to increase the likelihood of a summary judgment.