I believe the mechanism you found with SET is probably correct. You might generate a small amount of H2, but these reactions are not run under an atmosphere of hydrogen like in other metal reductions, so the hydrogen is probably just lost to the atmosphere.
Also remember that H2 doesn't act as the direct reductant in catalytic hydrogenation – it is likely a metal hydride or a hydrogen radical that does the reaction. These intermediates are also possible in nitro reduction, I suppose, but I don't know all the evidence for why the SET mechanism is preferred.