As a scientist, I often have to deal with similar BS of a reviewer misunderstanding a submitted journal article, literally insisting we do something we already did, and explained in the paper.
I try to just own the fact that I didn't make my explanation clear and visible enough, so I just submit an updated version along with a comment like "We thank the reviewer for pointing out that we didn't clearly articulate our work on X, so we revised the manuscript accordingly."
In this case change around the UI for this tool to make it slightly more idiot proof, resubmit, and it will be approved. Getting angry and adversarial will only make them have a bad attitude and intentionally be unhelpful. Take them seriously, stroke their egos, and take fully responsibility for having presented your feature in a way they didn't understand. Thank them for pointing out this huge flaw in your program and giving you the opportunity to fix it!