If you don't have it already, find a copy of Carroll Smith's book Nuts, Bolts, Fasteners & Plumbing Handbook. Though I'll never get as much out of it as someone with some engineering training or background, it is an oddly good read and I go back to it from time to time to research/relearn something. That torque wrench discussion has been one such push for me to go back to this book, as with new learning from that discussion I can likely better comprehend some of Smith's teaching.
To this specific topic, Smith comments that standard AN torque tables are generally compiled for plated fasteners without lubrication, whereas critical assembly fasteners generally have their own recommended torque values and lubricant (like the ARP bolts discussed). He also explains the order of magnitude difference that lubrication and plating can have on the ultimate impact of torque: "Just to give an indication of the percentages we're talking about, in the worst case, with dry and unplated threads, about 50% of the torque applied in tightening a common nut and bolt can be used up in overcoming the friction between the bearing surfaces and the work faces, while about 40% can be used up in overcoming thread friction and 10% actually contributes to preloading the bolt!"
He also explains why we "sneak up" on the final torque by going through a few intermediate steps, that the shear stress (that stress developed by the tightening friction) gets relieved when motion/rotation ceases - thus the final torque has less loss to friction and more put to preload.
