Some model based considerations on observing generation times for communicable diseases