Because we all know that looking at one, isolated, narrow FOV image proves the Earth is flat - even though we know from Visually Discerning the Curvature of the Earth and my post on high-altitude balloon footage that we require a wider field of view to even expect to see the curvature. This is how Flat Earth arguments 'work' -- for some odd reason, you have to look at only their evidence without considering all the evidence.
What do those other frames look like?
The numbers I get for 65 miles gives us a ground distance to the horizon of 712 miles (and viewer distance of 720 miles). With an apparent FOV of 60°/80° for each of the individual photos we get sagitta of 94.9/165.8 miles and a viewing angle for that much sagitta of just 0.1°/0.35° -- which would be just a few pixels per frame (my image is 419 pixels, divide by 80 to get pixels per degree, multiplied by 0.35 degrees = 1.8 pixels). You can also see how that 165.8 miles is almost completely on edge to the viewer because it is curving down and away from us, this makes it almost impossible to see any detail in that thin line of horizon.
As a reminder, this angle is the viewing angle from D to E below (Z to horizon₀ in the side view).
Which is what we find when we look carefully at frames where the horizon most closely passes near the center of the lens so they have the least amount of distortion:
The totally of the images seem to cover slightly more than 180° of view which means the stitched panorama is actually flattened out and should curve even more.
See Also: Air & Space: First Photo From Space