Comment on: Sequences of regressions and their independences