Some remarks about the scaling of systems of reactive Boltzmann equations