Why should there be much resistance to pulling the NT out from the bundle?
Why should it be dependent on the distance you have pulled?
I would expect that pulled tube will glide out from the bundle without much resistance, even if that would be a very tight configuration and that there may be small “bumps” when the atoms have to pass each other, but those would be rather small since that surface-surface interaction is very smooth.
Given the typical size of fluctuation of data from atomic scale simulations, particularly properties that are very sensitive like stress, I would expect that you need some extreme amount of averaging to get meaningful data that consistently becomes visible above the noise.