Inaccuracy due to lack of level will be small and can be accounted for. What isn't necessarily small and is difficult to account for is varying loads due to local winds.

I wonder if some of those with larger wheels were expected to operate from grass fields where the larger diameter offers lower rolling resistance and lower the bending load on the thinner/smaller nose gear. I note the balloons used on bush planes, for example. The mains, being already beefy would not require a lower rolling resistance.

The aircraft knows the wind direction and speed, right? I'd be interested in knowing if Boeing has determined the aerodynamic loads on the airframe for a stationary aircraft on the ground according to wind direction and intensity; and then in turn feeds this info onto the AUM / CG calculator. My guess is this is not accounted for but it's just a guess.

Not when it is parked on the ground and wind is turbulent over nearby buildings or shifting direction or isn't steady. They can make worst-case take-off and landing calculations based on the expected peak and direction, but there's a reason the windsock is on a swivel.