<p>Early-stage architectural design represents the most critical phase in affecting the performance of the lifecycle of a building. Although Multi-Objective Optimization and Building Performance Simulation tools have proliferated, their successful incorporation into the conceptual workflow remains a challenge. The systematic review critically investigates 226 scholarly works (2016–2025) to go beyond the survey of tools and examine how the structural interrelation of particular platforms determines the design space. The PRISMA 2020 guidelines were followed in order to synthesize the research via Scopus and Web of Science databases to answer the questions connected with the dominance of algorithms, integration of metrics, and research-practice barriers. The analysis shows that the most dominant one is a Common Stack with Evolutionary Algorithms (mainly NSGA-II) embedded into the Rhino/Grasshopper environment and linked to EnergyPlus/Radiance. Although the visual intuitiveness of this ecosystem is high, the resultant quantitative measure of a substantial Divide in the Geometry-System is based on the findings. Use of visual parametric tools establishes a pragmatic approach to geometric and facade level optimizations at the cost of system level variables (e.g., HVAC setpoints) that frequently demand code-based environments. Moreover, while the crucial criteria such as energy, thermal comfort, and daylighting are well-integrated, emergent metrics such as embodied carbon and indoor air quality remain critically underrepresented. In a bid to overcome the high computational cost and translation gap between architectural intention and mathematical construction, the literature is more inclined to use Machine Learning surrogate models and Multi-Criteria Decision-Making schemes. This review identifies a research roadmap of the necessity to have an accessible, low-code, and holistic performance-based approach to design that bridges the gap between computational capacity and professional practice.</p>