Dave,
Are we discussing rotor-propeller interaction, fuselage-propeller interaction, or wing-propeller interaction here? The objective of the discussion, hence the reason for my explanations, seems to be constantly moving.
The results in the cited NACA report do not even begin to suprise me. This is the difference between drawing in laminar air and turbulent air. Turbulent air is caused by the fuselage boundary layer. The counterrotating prop gains efficiency by not producing downstream swirl, but the penalty is the rear prop is already running in turbulent air off the front prop. Hence fuselage turbulence becomes insignificant, so pusher or puller makes no difference.
Non of this says anything about how rotor effects the prop inflow. You will notice that in both designs the prop is pushed as rearwards as possible so as to minimise the cyclic variation as main rotor blade passes overhead. Cyclic loading from uneven inflow distribution is the main cause of fatigue failure in propellers. A simplistic evaluation of streamlines just can't capture this, but it must too be considered.
Like i stated earlier pusher prop is a compromise, mostly chosen from deciding just how far into the unknown a new design should venture. Since X2 is gently accumulating flight time i would say they got it right...