A recent paper by Schall et al. (2020) concluded that beech forests managed in even-aged (EA) rotation systems were more efficient than unmanaged (UNM) forest and forest managed in uneven-aged (UEA) selective cutting systems in supporting landscape-scale biodiversity in Germany. The authors based their conclusion on a comprehensive multitaxon survey and a promising resampling model for assessing gamma diversity at landscape scale. Here, we challenge their conclusions and evaluate the importance of UNM forests for conservation of forest biodiversity. The average amount of dead wood reported from EA stands (27.8 m³/ha) was almost 30% higher than reported from UNM stands (21.6 m³/ha) in the study. Averages from long UNM temperate forests in Europe are typically six to seven times higher (131–157 m³/ha). We therefore conclude the UNM studied stands to reflect legacies of former management, and to be poorly representative of UNM forests. Data from our own studies, including long UNM beech stands in Denmark, demonstrate how this shortcoming seriously undermines the general validity of the presented results to conservation of forest biodiversity. Synthesis and applications. Preservation and restoration of intact forest ecosystems remains essential to biodiversity conservation. We show that the findings of Schall et al. (2020) do not contradict this important notion. Schall and colleagues identified UEA management systems as potentially inferior to more traditional EA management systems for conserving forest biodiversity at the landscape scale. The paper also provides insight into the limited short-term conservation value of simply abandoning forest management in intensively managed landscapes. Based on this, we call for discarding the current orthodox view of non-intervention when new forest reserves are created in temperate Europe. Active reinforcement of natural disturbance regimes and active habitat creation may lead to faster recovery of natural stand structure and forest biodiversity.