We are presenting in this work, the final design of the Multi-Object Adaptive Optics system for the Gemini InfraRed Multi Object Spectrograph (GIRMOS). This report outlines key modifications made during the critical design phase as we progress towards the manufacturing assembly integration and testing stage. Our adjustments include in particular, the wavefront sensor subassembly through the selection of a new camera and by optimizing the number of sub-apertures and pixels. Additionally, recent advancements in the Gemini North Adaptive Optics (GNAO) design prompt a reassessment of Ground Layer Adaptive Optics (GLAO) and tomographic performance, thus influencing both multi-Objects Adaptive Optics (MOAO) and the GIRMOS imager. The final optical and mechanical design of the MOAO subsystem as well as the revisited performance is discussed. Since the beginning of the project, we have identified some risks associated to the MOAO system and developed mitigation strategies and activities. We report the progress made towards the mitigation of identified risks such as open-loop calibration and control, go-to errors, quasi-static errors, etc. A primary objective driving the development of GIRMOS is the survey of a substantial sample of high-redshift galaxies. Using AO observations of z similar to 2 galaxies, as seen with SINFONI at the Very Large Telescope, and employing image processing techniques with various GIRMOS simulated point spread functions (PSFs), we methodically explore the GIRMOS ability to detect and characterize star-forming clumps within high-redshift galaxies. This analysis provides valuable insights into optimal target selection based on their positioning in the field of view, on the observing conditions (such as seeing, zenith angle, etc.), and the intended scientific objectives.