Plenty of "state of the art", at least a decade ago, that was not very state of anything. I think bias is inherent in our literature and solutions. But also, I

Plenty of "state of the art", at least a decade ago, that was not very state of anything. I think bias is inherent in our literature and solutions. But also, I agree that the probability of a better solution degrades over time (assuming that the implementations themselves do not degrade - building a faster hash table does not matter if you have made all operations exponentially more expensive for stupid, non-computational, reasons)