The deployment of traditional computing, control and monitoring paradigms in modern smart grids is characterized by several severe limitations. These could restrict their capability to manage the huge complexities and the vast penetration of distributed and renewable energy resources that are expected in future operational scenarios. In this context, the design of more scalable and more flexible solution paradigms represents a relevant issue to address. The enabling technologies and methodologies aimed at addressing these complex challenges include decentralized computing, self-organizing sensor networks, proactive control, and holistic computing frameworks. Accordingly, in this paper a selection from literature published on these topics will be analyzed and reviewed, in order to outline the most relevant research trends, and the main open problems.