A combined analysis of the Jefferson Lab data on nuclear transparency in A(e,e'π⁺) and A(e,e'K⁺) shows that the onset of color transparency (CT) is not universal across meson flavors. The pion transparency is well reproduced by the standard quantum diffusion model (QDM) with Δ M²_π ≃ 0.7 GeV², whereas the kaon data favor the quadratic expansion of the naive parton model (NPM) with the natural hadronic scale R_K ∼ √(σ_KN/π). This dichotomy cannot be repaired by physically motivated parameter choices: the pion slope excludes the quadratic expansion with any physical radius, and the kaon slope requires a QDM excitation scale far below the range conventionally used in pion transparency analyses. A microscopic interpretation, in which the diffusive evolution of the pion appears as an exceptional consequence of its Goldstone-boson nature while the kaon follows the generic ballistic expansion, is discussed.