Hard turning of hardened steels requires simultaneous control of surface integrity, thermal behavior, and productivity, which becomes more challenging under minimum quantity lubrication (MQL) due to its limited cooling capability. Although nanofluid-assisted MQL has demonstrated improved tribological performance, many machining optimization studies rely on surrogate models without rigorous statistical verification, potentially reducing the reliability of the obtained optimal solutions. This study proposes a statistically validated multi-objective optimization framework for hard turning of heat-treated AISI 4140 steel using coconut oil–based MQL enhanced with CuO nanoparticles. A Taguchi L27 experimental design was employed to evaluate the effects of cutting speed, feed rate, and depth of cut on surface roughness (Ra), cutting temperature (Tc), and material removal rate (MRR). Response surface models were systematically validated through analysis of variance, residual diagnostics, multicollinearity assessment, and prediction error analysis before optimization. The results reveal distinct governing mechanisms: Ra is primarily dominated by nonlinear feed effects associated with chip formation kinematics, whereas Tc is mainly controlled by depth of cut due to thermomechanical loading at the tool–chip interface. Multi-objective optimization using NSGA-II generates nominal Pareto trade-offs, and three selected decision regimes are locally confirmed through single-run machining tests. The proposed framework provides a statistically reliable decision-support approach for sustainable hard turning.