Remarks on classification of theories by their complete extensions