"Double evaluation" is the system setting and technical foundation of territorial spatial planning system. Based on the practical cases, this paper gives theoretical reflection on the Trial Guide. The results show that the scientific questions and hypotheses are constructed into the technological and logical paths of "double evaluation". Meanwhile, "double evaluation" exerts its impact by constraint and correction of planning system institutional construction. "Double evaluation" plays an effective role when it combines science and policy, but it is difficult for both sides to be perfect. This kind of problem on technical methodology and data aspects can be interpreted as "add-on dilemma", "decomposition dilemma" and "sorting dilemma". To achieve a reasonable and effective "double evaluation", it is necessary to improve the technical rationality and coupling of the evaluation content. Meanwhile, through the understanding of the difference between scientific research and decision support research, based on full adaptation to decision-making mechanisms and subject objectives, the evaluation method should be optimized according to the hierarchy and arrangement of governmental authorities. The conclusion of this paper is of practical significance to the position optimization and technical logic improvement of "double evaluation".