Engineering design optimization objectives span multiple dimensions, such as efficiency and sustainability, alongside advances in simulation accuracy, making better-informed decisions a priority in responding to industrial demand, where digital counterparts play a significant role [1]. For instance, simulation-based design optimization is often computationally demanding due to the need for high-accuracy simulations. Multifidelity approximations are regarded as promising tools that integrate exploration across many low-fidelity simulations with the accuracy of targeted exploitation of a few high-fidelity simulations, thereby enabling more accurate predictions in less time [2]. Since these methods utilise different fidelity levels, such as spatial discretization or the use of different physical models (Reynolds-Averaged Navier Stokes equations or potential flow), they enable substantial computational time savings. In this context, the present study assesses the effectiveness of principled fusion of multifidelity computational schemes that integrate active learning based on uncertainty quantification to inform the design optimization task of high-dimensional sea-domain vehicles. The optimization aims to minimize hydrodynamic resistance in calm water while constraining hull-form variations within specified limits to control displacement variability. The present study evaluates the performance of multifidelity frameworks developed based on different physical solvers as potential flow and Reynolds-Averaged Navier Stokes equations [3], and various spatial discretization levels as multiple levels of fidelity [4]. The discussion of the efficiency of multifidelity frameworks is extended to the effects of adaptive sampling strategies, geometry parametrisation, dimensionality, and the number of fidelity levels, providing insights for engineering applications that can serve as guidelines. Adopting multifidelity frameworks resulted in a significant reduction in calm-water resistance, and the efficiency gains are validated by high-fidelity solvers.