No, the error is not in the calculations, but in the measurement technique. All measurements are recommended to be carried out using the Fine aiming line, which is activated by pressing this button.
When measuring, it is necessary to capture the extreme row of pixels of the lower and upper extremities of the solar disk.
In addition, measurements of the diameter of the Sun should be carried out on a pause. Otherwise, between the first and second measurements, the sun can change its altitude by a certain amount, the profile of the solar disk (outlines, set of pixels) also changes by crosshair movement - this is purely "computer reason" introduces an additional error of measurement. You can check your measurement method yourself: take 5 measurements - if the variation is more than 0.3 arc minutes, then the quality of the measurements is poor.
In my series of 10 Sun measurements, the variation (difference between Max and Min result) was 0.2 arc minutes. 1.61666666 and 1.61333333. I took the average of 10 measurements as the result.