Go do the statistics on that to explain why that's reality.
The issue has nothing to do with statistics, the problem is that there are two unjustifed assumptions in the hypothetical provided.
First, a call plot from a dryfire session does not tell the complete story as far shooter accuracy is concerned.
Second, the free-recoiling accuracy of a rifle does not fully characterize its performance from all shooting positions.
I responded to a comment about the combination of a 1MOA shooter and a 2MOA rifle. My response was based on the very reasonable assumption that those numbers were actually representative of the live-fire capability of the two systems and that the testing used to generate those representative figures (including the combined figure) is done under conditions that are reasonably similar.
If one has truly representative accuracy figures for the two systems and those two figures are 2MOA and 1MOA, then the resultant combination for the two systems, under similar testing conditions is going to be about 2.24MOA.
Nothing I've said was intended, nor should it be construed, to mean that it's impossible to construct a scenario whereby it's possible to combine a 1MOA shooter and a 2MOA rifle and get groups that are larger than 2.24MOA. That is certainly possible.