On a Problem of Repeated Measurement Design with Treatment Additivity