NOx from big industrial processes tends not to be an issue. Catalytically reducing it to N2 and O2 is easy, and at the same time, you get out 'free' high grade heat, which there is usually some use for elsewhere in the plant.

> Interestingly, the paper mentions that the reactions do produce (some) hydrogen, which could perhaps be recaptured.

The hydrogen is used for reducing the iron oxides, that's the whole point of the process! That is,

FeXOY + H2 => Fe + H2O