Your main point might be real, but until we manage to make it work around 0K, i'd have a hard time to believe a carnot engine can have an efficiency close to on

Your main point might be real, but until we manage to make it work around 0K, i'd have a hard time to believe a carnot engine can have an efficiency close to one. I think current "future" tech aim for something like ~50% (theorical) efficiency, and that's with a _very_ wide temperature differential that would put a real strain on materials.