Quote:It matters because the measurement was at close focus, so the effect might seem worse there than in "reality"?
See the sample images. The borders are softer than the MTFs suggest.
I think you should get over the idea that the measurements are close focus.
The object magnification is FAR beyond anything that you would consider to be a close focus scenario. Close focus ends a 1:10. At some stage I calculated this - was it 1:50 ? Something like this.
The object magnification is the same across all tested lenses.
In any case it just doesn't matter here - the field curvature is visible at infinity.
If I didn't spot this in the sample images I wouldn't have reported it (plus/minus human error that can always occur even during measurements). The MTFs are the beginning of wisdom but not the end of it.