It is important to save the time to have information about the moment of time that saved quantity samples belong to. If step-size is fixed you can calculate time position from position of samples in a saved array, as (position-1)*Ts, but if step-size is variable (when you simulate continuous systems with auto step-size) you can not recover time information from position alone, so the time is additional signal that is save. Saving time makes thing much easier even if sample-size is fixed.