IF the stress loadings are that close on the fasteners that anti-seeze dangerously effects that, then I say the bolt is underengineered for the application.
If by "stress loadings" you are referring to the difference in the turning effort during torquing and resultant increase in clamping load, greater tension load and stretch value applied to the fastener?; that is all likely true but that is not as related to me the real concern.
Rather, as relayed to me repeatedly, it's a simple concern for the very reason one wants to apply anti-seize, a reduced effort in the removal of said fastener! I think this opinion probably arose with the earlier use of the press-formed steel wheels which tend to flex, perhaps more than the engineering intentions (later renditions are superior
), this tending to 'work' the fasteners attempting to maintain the clamping of the wheel to the hub.
Evidence of this is witnessed with the fatigue cracking often in close proximity of or directly emanating from the lug holes of the wheels. This even more so with the introduction of the radial tire which caused addition flexing of the rim, and was addressed with redesign for them once it became apparent.
Scott.