My personal feeling on precision (not accuracy - still waiting for you to figure out what "accuracy" means)...
...are that in a single core, of course the accuracy is very poor, as stated. But (just like where you can add multiple images together to get more information than the number of pixels captured in each individual image) when you add the information from many many cores to your dataset, you can get more precision from the system than any one core could provide. I believe that your spokesman (who I take to be an intellegent scientist who knows better) was deliberately misleading.