Real-time computer graphics technologies such as graphics engines and graphics pipeline (software and hardware components) have improved considerably in the past decade. However, increase in efficiency and broadening of the applicability area has come at the cost of complexity of the tools. Therefore, development costs of advanced visualization systems can increase considerably. In this research we explore the possibility of addressing this problem by analyzing the software development methodology of graphics pipeline. We believe that the pipeline's limitation regarding geometrical primitives in its operations is one of the reasons for the development efficiency bottleneck. We propose an approach that extends graphics pipeline with additional processing stages that can operate with primitives of higher order. This new method has the potential to reduce the pipeline's application-level complexity for developers to create better software.