Yeah, I think test-of-time awards are a better indicator of which papers actually had solid impact; there's often very little overlap b/w best paper awards and test-of-time awards.