Over the past four decades, the mobile communications industry has evolved from the first generation (1G) to the 5G premises, while the research community is currently discussing the broad outlines of the 6G system foreseen from 2030 onwards. In fact, even if 5G is extensively advocated as the expected ecosystem to fully meet the requirements of internet of things (IoT) applications, e.g., autonomous robots and vehicles, extended reality (XR) platforms, and brain-computer interface (BCI), it is only by the advent of 6G that such a bold achievement can be truly met. One key milestone towards the 6G vision is the development of a multi-objective analytical tool enabling the joint optimization of communication, control, localization, sensing, energy, and other parameters and resources. In this paper, we discuss how stochastic geometry (SG) can serve as a potential candidate for such aim. We also outline some fundamental limitations of SG, as well as some avenues to address them, and develop new seminal results ready-to-use in the analysis and modeling of future 6G networks.