//Unrelated conversation below.
I will have to disagree with this. While the torque rating of the fastener depends on how large it is and what materials it is made of. The pressure between a bolt head and mating surface is provided by the bolt threads. If you were to zoom in on them at a microscopic scale you would see that the more you torque the bolt the greater the threads deform. Much like if I were to hold one end of a metal rod in a vise, and push down on the opposite end. The reaction of the rod is to provide an upward force on my hand, but at some point the strength of the metal is overcome and the rod (thread) breaks.
Without friction the bolt would simply loosen itself when you let go. The friction force between two surfaces is directly proportional to the perpendicular force between them and the coefficient of friction. Even if anti-seize is not primarily a lubricant I am nearly certain it does have lubricating qualities. A lubricant lowers the coefficient of friction.
well, you are kinda saying the same thing I was saying with the first paragraph there, but, with lugnuts, your calculations for friction are in a slightly different direction because of the tapered ends. This kind of throws off some of that stuff. This means you have more surfaces that you have to calculate friction against, and the force distribution is different on the stud. Also, you have to calculate the spring rate of the surface it is mounted against and that of the threads.
Either way, you are looking to stretch or deform the threads ever so slightly when torquing a bolt, so that is what generates the friction. The coefficient of friction of the actual surface does not matter as much because this force multiplies the friction by so many times that small changes in surface friction don't really matter.
Still, while anti-seize does have lubricating qualities, you still generally want that low friction when torquing things. Like I said before, you oil head bolts before you install them.... The holding force is coming from the threads, but it is because of the pressure on the threads, not so much slight changes in the coefficient of friction of the surface.
I mean, when it all boils down to it, you are applying a ton of force to a very small surface and you are talking about relatively small changes in the friction coefficient. It would be like saying that a lug nut on a new OEM wheel stud has more chance of untorquing and falling off than on an old rusted wheel stud after you install new nuts, when in fact, the opposite is true.