In the redox titration of $\ce{FeSO4}$ with $\ce{KMnO4}$, the colour change of the solution at the end point is colourless to light pink. Most laboratory manuals and books advice us to check whether the light pink colour stays or not for 30 seconds. If the colour stays after 30 seconds, then end point is reached. If the colour disappears before 30 seconds, then the titration is not finished, so more $\ce{KMnO4}$ should be added.
As far as I understand, at the end point, all the reducing agent is depleted. So, the colour of $\ce{KMnO4}$ stays. And, if I have interpreted my book correctly, the colour of $\ce{KMnO4}$ is to stay for more than 30 seconds so that we can say that the end point is reached.
This means, that the colour of $\ce{KMnO4}$ might disappear after 30 seconds even though the end point is reached. But why? All reducing agents are depleted. Aren't they?
Now, I have found some reactions which might possibly cause the discharge of colour after the end point:
\begin{align} \ce{3 MnSO4 + 2 KMnO4 + 2 H2O &-> 5 MnO2 + K2SO4 + 2 H2SO4}\\ \ce{3 MnSO4 + 2 KMnO4 + 8 H2SO4 &-> 5 Mn(SO4)2 + K2SO4 + 8 H2O}\\ \ce{4 KMnO4 + 2 H2O &-> 4 MnO2 + 4 KOH + 3 O2} \end{align}
I am not sure which of them occur in reality. So, which of the reactions are the reasons for the 30 second rule? And why is it 30 second anyway? Why not 40 second or 1 minute?