That source only covers losses due to iron evaporation during combustion, it doesn't cover reduction losses back into iron or what the regenerated iron's particle size is, which tells us what milling energy is needed to get back down to sub micron iron. That's the inefficient part of the energy loop, along with reduction efficiency with green hydrogen.

You need to understand that whatever method is being used for reducing your powder back into iron metal is going to be a mature technology already used for steelmaking. If green energy scientists somehow found a new method for reducing iron into metal it would be global breaking news in the steelmaking industry. That's the extraordinary claim you need to support with sources: how it gets regenerated and at what efficiency, does it need to be milled and to what starting and finishing particle size, consuming what energy, CAPEX, and OPEX? What's the total loop efficiency?

I really think the overall efficiency is going to be terribly low once you do the math on all steps in the cycle...