I had a chance to perform a randomized study of stapling today. My results suggest that staples parallel to the top edge (A) and placed a random distance from the top results in the most efficient paper stacking. Staples parallel to the side (B) tend to be placed more conservatively towards the edge with a reduced variation of placement and thus an increased stacking depth at staple overlap. Diagonal stapling (C) has less variation in placement but more variation as expressed in minutes of rotation from forty-five degrees resulting in consistently the largest paper stack thickness.

Sample size:
n of (A) = 14
n of (B) = 0
n of (C) = 0

Studies of friction in each direction were not studied to determine mean proportion of stack slide.

On an unrelated note, the coefficient of friction of a corn cob is somewhere around 0.83154