Modern power generation units require highly efficient control systems due to increasing sustainability demands. Traditional methods for evaluating control quality mainly rely on technical indicators, which inadequately reflect the economic consequences of control deviations. This study presents a unified methodology for assessing control quality based on economic efficiency criteria for both deterministic and stochastic processes. The approach incorporates integral quadratic criteria and statistical moments to quantify the economic impact of deviations in key controlled variables, such as steam temperature, pressure, and air excess coefficient. A procedure is proposed for determining the exponent that characterizes the sensitivity of economic indicators to deviations in control variables. Both graphical and analytical methods of evaluation are presented. Furthermore, an analytical relationship between generalized and classical quality indicators—variance and quadratic criterion—is established. The methodology was applied to industrial data from 300 MW power units at the Zmiiv Thermal Power Plant under the regulation of air excess coefficient and steam pressure. The results show that the proposed criteria provide a more comprehensive and economically meaningful assessment of control performance compared to traditional technical indicators. The proposed approach represents a practical tool for improving the economic efficiency of automatic control systems in thermal power engineering and contributes to the achievement of sustainability goals.