v1.2.0 headline. A CONUS 48-state x 4-scenario sweep at the 50-year horizon under canonical B1.3 FIA EXPNS inventory. Under the historical disturbance schedule (HIST), CONUS shifts from a +4,870 TgC ecosystem sink (BAU no-disturbance counterfactual) to a -10,109 TgC source (15,000 TgC swing). Reduced harvest (RH, -30% probability) recovers +3,001 TgC vs HIST. Reduced clearcut (RH_CC, -50% clearcut, partial unchanged) recovers +5,042 TgC, the most efficient policy lever on a per-area basis. Climate-amplified wildfire (WARM_HIST, region-specific multipliers) adds another -998 TgC at CONUS scale. 36 of 48 states are sources under HIST vs 7 under BAU. The Pacific Northwest is the most extreme regional source under every scenario (-80 to -90 Mg/ha at 50 years).Scientific finding (v1.0). Two implementations of the CBM-CFS3 forest carbon model family, libcbm (Canadian Forest Service Python/C++ reimplementation) and GCBM (the moja FLINT spatially explicit engine), are compared across six US states under matched parameter conventions. The year-five carbon density gap is approximately +24%, with libcbm/GCBM ratios between 0.74 and 0.80 in WA, MN, IN, ME, OR, and a single warm-donor outlier in GA at 1.05. Replacing the uniform forest-type stratification used in legacy parity runs with a stratification anchored on FIA EXPNS expansion factors (the canonical FIA Total Area Estimator) adds +6,760 TgC to the CONUS year-five carbon stock at n=39 states, a +14% shift on 246.5 Mha. Inventory stratification, not engine implementation, dominates cross-model uncertainty for this generation of CBM-CFS3 carbon models.Regional DOM-pool fingerprint. Mean libcbm slow soil carbon under B1.3 FIA EXPNS is 184 Mg/ha in the PNW, 143 in the Atlantic Maritime NE, and 105 to 119 Mg/ha across the South, Lake States, and Mountain West. A Q10 mean annual temperature sweep on Oregon (MAT 4 to 13 C) shows the slow-soil overshoot is established during spinup rather than emerging from runtime Q10 scaling.Contents v1.2.0. 48-state libcbm year-five pool outputs (B1.3 FIA EXPNS); 48-state libcbm 50-year baseline trajectories (BAU); 192 50-year scenario trajectories (HIST, RH, RH_CC, WARM_HIST across 48 states; bundled as one tarball); CONUS, regional, and per-state scenario rollup CSVs; 6-state libcbm-vs-GCBM density gap matrices under both B1.1 v6 parity and B1.3 FIA EXPNS; per-pool and per-DOM-pool decompositions; F3 Q10 MAT sensitivity sweeps for Georgia and Oregon; seven publication figures (Fig 1 through 7 = CONUS scenario sweep panel); and the cbm_conus v0.2.1 source tarball.Pipeline. All libcbm outputs were produced with the GCBM2hpc pipeline on OSC Cardinal (allocation PUOM0008). Scenarios were generated with the cbm_conus tools/raster_to_sit_events.py in dry-run mode (literature-calibrated regional disturbance rates; production path reads HCS phase 5 + TREEMAP probability rasters). Six-state GCBM aggregates were produced spatially via moja FLINT containerized GCBM at 1-degree WGS84 tiles. The CONUS economic and ecosystem services framework that consumes the scenario sweep is at github.com/holoros/cbm_conus.Companion materials: methods note at holoros.github.io/perseus-forest-intelligence/methods/inventory-stratification/ and American Forests comparison dashboard at holoros.github.io/perseus-forest-intelligence/methods/af-comparison/.