The latest advancements in glazing technology are driving facade design towards complex and adaptive fenestration systems. Accurate simulation of their optical properties and operational controls for building daylight performance evaluation requires advanced modelling techniques, such as climate-based daylight modelling (CBDM). At the same time, computational efficiency is key to quickly simulate this complex performance over a full year. Over the years, several CBDM techniques were developed to answer these two main challenges, but they were never systematically benchmarked against each other. This paper compares state-of-the-art Radiance-based simulation techniques in terms of annual daylight performance metrics required by national guidelines and international green building rating schemes. The comparison is performed on three different shading systems: diffuse Venetian blinds, specular Venetian blinds, and perforated solar screens. Findings show that simulation methods are characterised by significant differences in their implementation and visual rendering, but most annual daylight metrics result in consistent values (within ± 20%). A notable exception is Annual Sunlight Exposure, which is highly sensitive to the chosen simulation method, with differences of up to 47 percentage points. Additional outcomes from the present work are used to compile a list of generalised recommendations for designers and policy makers.