Assessments of marine carbon dioxide removal (mCDR) often overemphasize biogeophysical capacity while underestimating deployment constraints, creating a "scale illusion" that undermines comparability across pathways. To address this, we propose "Triple REAL", a screening framework that shifts the evaluation focus from theoretical capacity to permit-bounded, MRV-credible throughput. Triple REAL comprises four deployability filters—3R, 3E, 3A, and 3L: (3R) realistic attribution under explicit boundary conditions; (3E) ecosystem- and equity-relevant risk control; (3A) actionable scaling grounded in infrastructure and time windows; and (3L) legal readiness under existing governance regimes. Applying this screen discriminates between open-ocean interventions and operationally bound semi-managed settings where iterative management is feasible. Wastewater treatment plant ocean alkalinity enhancement (WWTP-OAE) emerges as a deployable point-source archetype constrained by verifiable permits, with wastewater alkalinity preservation (WAP) occupying a comparable wastewater-linked range. Under the SSP2-4.5 scenario, this could yield a cumulative removal of 2.7–24.3 GtCO₂ by 2100 coupled with gray carbon, while wastewater alkalinity preservation occupies a comparable wastewater-linked point-source range Furthermore, the BCMS macroalgae-farming case demonstrates how measurable non-biomass carbon pools support MRV while adhering to risk-bounded intensification, offering an additional 14%–29% carbon sequestration capacity relative to harvested yield through integrated organic and inorganic burial. Overall, Triple REAL facilitates a transparent classification of mCDR proposals, prioritizing pathways that deliver verifiable net removal under real-world constraints.