The extrapolation of the original cosmic ray primary spectrum derived from the size spectrum measured in the Akeno classical EAS array coincides with the spectrum measured recently by the HiRes Stereo experiment. After revisiting the calibrations carried in the overlapping energy region around 1018 eV, we discuss the consequences of the different approaches in classical arrays and giant surface arrays. The data are obtained from the size spectrum registration in the Akeno experiment with a modest space grid of 30–100 m, instead of 1 km or more in giant arrays using density estimators in place of size and different absorption treatments for inclined cascades. While the analysis of those circumstances suggests a reduction of the energy converted from the estimators in giant surface arrays, the consistency of classical and fluorescence measurements gives more support to the GZK prediction.