Look at the tank with the letters 'AGR' on it. Look at the jagged edges. Those are not on the film originals. The scanning removed information, homogenizing fine detail into single-tone squares. It's not like the full image information from the film is somehow "compressed" into the smaller format's scale.
How can you expect a roughly 35 micron pixel to match a 5.2 micron pixel? It might seem that a 6 times larger pixel of the same scene on a format about 5X larger on the short side would make for a good comparison, but it does not, because the film resolves much smaller than 35 microns.
The fault is in trying to shrink the 8x10 to "match" the smaller digital through a low-res scan. A correct test would have enlarged the digital to match the 8X10. A comparison of more or less equal file sizes is meaningless, because full-information files are not equal-sized. Comparing equal-sized full-information prints from the originals would have been meaningful. Pixel-peeping one full-information image against one part-information image proves nothing.