This paper discusses the value proposition of a system-level Digital Twin (DT) in the context of complex manufacturing processes from the managerial perspective. The central promise of DT-technology is to use and iterate the available real-time process information in a simulation setting transforming it instantly to operational or managerial-level decision-making implications. Despite the clear potential of this emerging technology, a gap of knowledge exists on how such DT could be implemented and what would be its defining features. The key contribution of this research is to lay out the central discrepancies between the promise of digital twin technology vision versus what is possible within the limits of current industrial infrastructure in the short- and mid-term. This research builds on the currently existing scientific literature which we use to point out ten foundational issues of system-level digital twins that are analyzed and discussed in detail. As a conclusion, we propose that large system-level DT-projects have a managerial rationale only when several preliminary conditions are met and fulfilled.