This paper investigates the performance and optimization of the Geometric Wavefront Sensor (GWFS) in openloop wavefront sensing, along with the Curvature Wavefront Sensor (CWFS) and Shack-Hartmann Wavefront Sensor (SH-WFS). The GWFS uses a ray tracing process to calculate the displacement of intensity fluctuations from two defocused point source images. While similar to the CWFS, initial simulations and testing have shown the GWFS to have superior performance compared to the CWFS. Various parameters within the GWFS -such as the signal-to-noise ratio (SNR) sensitivity, the number of Radon angles, the virtual propagation distance, and the number of reconstruction modes -are explored on a laboratory test bench. We found that the GWFS wavefront estimate error experiences an inverse relationship to the SNR, a minimum of 5 Radon angles is required to accurately estimate the single Zernike mode wavefronts (Z 4 -Z 15 ), the virtual propagation distance is confined by ray crossing and Fresnel diffraction effects, and the number of reconstruction Zernike modes is limited by noise amplification and over-fitting. This paper demonstrates the capabilities of the GWFS, illustrates the resulting wavefront estimates, and confirms the superior performance of the GWFS compared to the CWFS. The optimized GWFS will be utilized at Mt. John University Observatory (MJUO) in New Zealand for satellite and space debris imaging and tracking.