Nowadays, model composition plays a central role on software engineering activities. For example, reconciling design models developed in parallel by different software development teams. It can be defined as a set of activities that should be performed over two input models, Ma and Mb, to produce an output composed model, Mcm. Usually, the problem is that the Mcm does not correspond to the output intended model. It can be explained by the fact that MA and MB conflict in some way, and the emerging conflicts must be detected and solved. However, these two activities tend to be critical as model composition is usually a cumbersome manual task. This paper presents the replication of an exploratory evaluation about the impact of aspect-orientation properties, such as quantification and obliviousness, on conflict resolution effort. The intuition is that improved modularity of aspect-oriented models may help to better localize composition conflicts. We investigate this issue through an application different from the one used in the previous study, in which three well-established composition algorithms (i.e., override, merge and union) were used to evolve the design models. The goal was to analyze how the impact of varying quantification and obliviousness degrees are correlated with the rate of semantic and design conflicts. We identified some cases where quantification and obliviousness reduce conflict rate and the composition effort.