For GDE scale-up, what fails first?

Opening a focused GDE thread because the lab-to-pilot jump still feels like the most useful place for hard-won experience.

Burdyny and Smith made the case for commercial-condition testing early (doi:10.1039/c8ee03134g). Wakerley et al. cover reactors and metrics (doi:10.1038/s41560-021-00973-9). The newer materials review is good for the gigatonne-scale framing: doi:10.1038/s41578-024-00696-9.

For people who have touched real cells: is the first wall catalyst coating uniformity, flooding, salt management, carbonate crossover, membrane durability, or something less visible?