On stationary optimal strategies for team LQG control problems