A promising quantum computing architecture comprises modules of superconducting quantum processors linked via optical channels using quantum transducers. As quantum transducer hardware improves, a need has arisen to understand the quantitative relationship between transducer-device characteristics and the strength of the resulting remote entanglement. Using Monte Carlo simulations that incorporate 2-to-1 and 3-to-1 entanglement distillation methods, our model maps transducer device performance up to system-level channel performance, thereby allowing the performance of remote entanglement approaches to be compared and optimized. We find the extreme photon loss (EPL) distillation protocol to be particularly high performing. Moreover, even without distillation, present-day transducers with added noise of photons are at the threshold of enabling remote Bell pairs with fidelities exceeding 50%. If the next generation of transducers can improve by 3 orders of magnitude in added noise, efficiency, and repetition rates, then they would allow for remote two-qubit gates achieving 99.7% fidelities at MHz rates. These results set practical targets for transducers to be ready for deployment into modular quantum computing systems.