（ Compared to the simulation results obtained using the Static Position Function in TIMSS, the results are approximately 1 mT lower. ）
I think you need to account for the actual sensor location in the package when placing your DRV5055. The sensing element is located in the center of the XY plane, but is elevated from the bottom of the package by 0.65mm in the Z direction. I matched your result for the first case with the magnet center at 5.5mm in Z and the point in space for the sensor at the origin.
I then created a linear function with the same magnet in the same vertical displacement moving from +3mm to -3mm in the X direction. I placed the sensor at (0,0,-0.65). Here is the resulting field at the center position
