Increasing renewable generation raises short-term forecast uncertainty while reducing the time available for congestion-management actions. This study investigates whether recent measurements can improve transmission-line loading forecasts during the final hours before delivery. Historical day-ahead, intraday, and real-time data from the TransnetBW control area and its surrounding network are analysed. A bias-corrected intra-day congestion forecast (IDCF) serves as a benchmark for a Continuous IDCF Trajectory Reconstruction and Real-Time Calibration method. The proposed approach reconstructs the forecast horizons of a single operational IDCF run as a continuous trajectory and calibrates its level using the latest real-time system state. Results show that real-time information substantially improves close-to-real-time congestion observability. The calibrated trajectory outperforms the IDCF as early as 45 minutes before delivery and achieves a 37,92% lower mean absolute error (MAE) than the final operational IDCF available at delivery time. These findings indicate that the primary value of real-time measurements lies not in describing the network state at delivery more accurately, but in providing actionable information before real time. The proposed methodology therefore contributes to reducing the remaining Close-to-Real-Time (C2RT) observability gap and supports more effective congestion management and system operation.