This paper looks into various aspects brought to light by numerical work on the generalized interacting winds model for planetary nebulae. First, a detailed comparison between radiative and non-radiative models is made, showing that one's naive expectations of the effects of radiative heating and cooling are not always true. Secondly, we consider the evolution of the slow wind after it has become ionized. It is found that the initial aspherical density distribution is smoothed out after ionization, thus requiring the aspherical nebula to be formed within the first few thousand years of PN evolution. Thirdly, the nature of the flow in the hot bubble is investigated. Both one- and two-dimensional numerical models show strong signs of instabilities or turbulent flow in the hot bubble. Although observationally hard to prove or disprove, this turbulent structure is critically examined. It is found that, although the turbulence is not a numerical artefact, the full three-dimensional picture will most definitely differ from what is found in two dimensions. The implications for the interpretation of the models are discussed. Finally, the issue of the soft X-ray emission from PNe is considered. It is found that soft X-rays originate mainly from the thin interface between the hot bubble and the actual nebula.