With the purpose of enabling a low-carbon future, power systems worldwide are undergoing major transformations. These developments require new advanced control and operation approaches to ensure a stable and efficient system operation. Distributed consensus-based algorithms are a promising option to provide the necessary flexibility and scalability to cope with these challenges and have, thus, been widely investigated in the literature. Yet, most available results are limited to scenarios with reduced-order models and ideal communication. Motivated by this, we perform a case study using a detailed dynamic model of the well-known Nordic test system equipped with a consensus-based distributed secondary frequency controller. Our main objectives are to analyse the robustness of the closed-loop system with respect to unmodelled (voltage and higher-order generator) dynamics as well as communication delays. To facilitate the later property, we employ robust-stability conditions in the control design. Then, the performance of the proposed controller is assessed through detailed dynamic simulations covering several disturbances leading to large frequency and voltage excursions.