Yes, this is a technical achievement, and may provide a model for studying parallel processing. It is, however, unfortunate that the media presents technologies like this and the memristor is as a models of the human brain since these do not resemble the biological form in function, structure, or processing result.
It will be wonderful if artifical sensing organs are developed for less fortunate handicap people. They will get new opportunities with enhanced senses. We study neural network for long. It looks, now time has come to see its application in real world.
How do these chips compare to biology in terms of density and clock speed? I cringe every time I see traditional computers compared to the human brain in terms of capacity because their architecture is so different from our meat machines, but this seems to be one that is more directly comparable. Also, how much visibility do researchers have into the operation of the device? I can see this as a real tool for studying emergent intelligence.