The authors have already developed and studied the Two-Geometry Roundabouts in terms of design principles. This paper in fact is the natural extension a previous article by Pratelli et al., 2022, in which the authors discussed the main concepts of the Two-Geometry Roundabouts and the details that determine their success. However, when someone talks about roundabouts in general, it is also essential to specify which respective capacity model is most suitable. The Highway Capacity Manual (HCM6th) provides a popular method for estimating the capacity of conventional roundabouts. Nevertheless, the Two-Geometry Roundabouts are classified as unconventional roundabouts, and therefore this paper aims to answer the following crucial question: “Which capacity model should be used to calculate a Two-Geometry Roundabout?”. Firstly, the impositions of the Italian guidelines, where all roundabouts with a non-circular shape or with an inscribed diameter greater than 50 meters are defined as unconventional, are described. The achieved empirical results in respect to some test has compared with those otherwise obtained through the HCM6th methodology. Next, two popular methods suitable for calculating the capacity of Two-Geometry Roundabouts are resumed, namely the SETRA method and the Brilon-Wu method. These two are chosen because on the one hand SETRA directly requests dimensions of some selected geometric components; while on the other hand, the Brilon-Wu approach is based only on a few macro characteristics such as the number of entry/circulation lanes and the values of the gap acceptance parameters. Then, the successive section is devoted to comparing performance between a Traditional circular Roundabout and a Two-Geometry Roundabout, designed for the same intersection. This comparison was carried out through the use of experimental results obtained with a dynamic simulation approach made with Aimsun software. Finally, the paper ends with a brief discussion on some possible insights framed in a further development of the research, which could involve the safety models of Two-Geometry roundabouts, and their possible extension to roundabout corridors.