Comment on “Struggles with survey weighting and regression modeling”