Escubic;
Very fine, skillful mathematical & graphic work indeed...tx.
We may see that the 'S' turns aren't nearly as violent as might be assumed from the fore-shortened view as taken by a strong telephoto lens. Also we may see that it doesn't take a large difference between heading and track to place the nosewheel outside the physical paths of the main gears. Such wide lateral "gyrations", (now far less so), while the nosewheel's track was straight, are realistic when considered as shown in this corrected photo.
I was searching all over for a program that would "know" that the perspective was in the original photo of the ground tracks and of course no program does - they only know pixels - it is us who interpret according to the rules of perspective...thanks for your excellent post.
Thanks to
PJ2 for his assessment which is what I support as well. Accepting the foreshortening effect of the camera lens it now makes quite plausible even low side forces applied near the aircraft centerline such as rudder and/or thrust asymetry to explain the tire tracks.