Remarks on precision
From 2.1 follows that neither the difference in size of sun and earth is
considered nor the effect of lightrefraction in the athmosphere.
Defined as sunrise and sunset are those moments when the center of the
sun appears at the horizon. The real daylength is much greater because
of the lightrefraction espacially in areas with low suns orbit. (Time
of dawn obviously is part of "night".) Therefore the uncorrectness
following the deviation from spherizity of earth is negligible (about 0,5%).
Earth is calculated as perfect sphere.
In the opposite it is considered that according to Kepler's laws the orbit of earth is elliptic what makes the velocity of rotation not beeing constant. This effect is so important that in northeren hemisphere winter is a complete week shorter than summer.
This effect is considered only approximately and not astronomically correct . Because of the fact that it's only a schematical visualization it should be sufficiant that in every year (not beeing a gapyear) the astronomical changes of the seasons will be the "usual" days. In order to calculate correctely you would have to consider exact initial condition for every special year.