0.07 acres off sound like a rounding error. I don't think it is the same problem.
Shape.Area is usually almost meaningless. It is the cartesian area calculated in the coordinates of the dataset as if the the world was flat. If your dataset's spatial reference is not defined to start with, the shape.area is of unknown units, so potentially doubly meaningless. Just think, what does the shape.area mean if the polygons are in geographic coordinates (latitude/longitude)? It means pretty much nothing at all.
If your dataset is projected with non-angular units (metres, feet, etc.) then it represents the area within the tolerance of the distortion factor (scale factor) of the projection system if your projection system is appropriate for the data and area measurements. In that case you may end up with 0.07 discrepancies due to projection distortion.