The accurate simulation of complex dynamic phenomena requires the availability of advanced constitutive models capable of simulating a wide range of features of soil behaviour under cyclic loading. One possible strategy is to improve the capabilities of existing bounding surface plasticity models, as this framework is characterised by its modularity and flexibility. As a result, specific components of the formulation of this type of model may be adjusted to improve the reproduction of any aspect of soil behaviour deemed essential to the problem being analysed. In this paper, a series of computational studies are performed in order to establish the impact of expanding a bounding surface plasticity model for sands on its modelling capabilities and to suggest ways of mitigating the associated increase in complexity. Changes to three distinct aspects of the selected constitutive model are examined: the shape of the Critical State Line in p′ − e space, the expression used for calculating the hardening modulus and the form of the yield surface. It is shown that the introduced changes have the potential to increase significantly the ability to control how certain aspects of soil response, such as degradation of stiffness and flow liquefaction with limited deformation, are reproduced by the model. Moreover, this paper presents a systematic approach to the expansion of this type of constitutive model, establishing how alterations to the formulation of a model may be assessed in terms of improved accuracy and potential benefits.