Coupling architectures — the infrastructural, contractual, and institutional arrangements through which a polity receives critical external flows — are not neutral conduits. They are attack surfaces. This paper formalises one case of the general mechanism: when a polity reduces dependency on one dominant partner by rapidly increasing dependency on an architecturally similar alternative, it has performed substitution, not diversification. Substitution leaves the maximum leverage term in the coupling portfolio invariant and introduces a predictable interval — the weaponization lag — during which switching costs accrue and political alignment drifts. The framework is applied to three 2026 cases: Iran's dollar exposure, Cuba's energy architecture, and the European energy reconfiguration of 2022–2023. The paper commits to five publication-dated forward predictions whose resolution postdates publication. This paper is a standalone application within the broader Coupling Geometry programme. The programme's axiomatic foundation and civilisational empirical base are presented in separate companion papers at https://doi.org/10.5281/zenodo.20137424 and https://doi.org/10.5281/zenodo.20137863. Manuscript under review at the Review of Radical Political Economy.
Philip Pepper (Tue,) studied this question.