As magnetic-confinement nuclear fusion research advances toward reactor-stage operation, dynamic control of plasma behaviour has become an urgent topic. In this context, heat transport models that are both accurate and computationally efficient will be essential for designing control strategies in future fusion devices. In magnetically confined plasmas, several mechanisms contribute to radial heat transport, but turbulence driven by microinstabilities — such as trapped-electron modes (TEM), ion-temperature-gradient (ITG) modes, and electron-temperature-gradient (ETG) modes — plays the dominant role. Because the underlying physics is highly complex, reduced advection–diffusion (AD) models with effective transport coefficients are widely employed, often calibrated against experiments. However, these closures rely on strong assumptions (e.g., linearity, quasi-stationarity, and locality) that may, under certain conditions, be significantly violated by the intrinsically nonlinear and nonlocal nature of turbulent transport.