A quick disclaimer: I know that I don't know enough about either chemistry or fluid dynamics (hence why I asked the question), so if you spot a mistake please correct it.
Assumptions: The air is at STP and is composed entirely of elemental nitrogen. I'm also assuming that the molecular binding energy is 0, since it represents such a small number compared to the others.
The total amount of energy present in either material is going to be:
Erest+Enuclear binding = Etotal
where nuclear and molecular are both binding energies. Let's just take a first order approximation of these energies to check if the binding energy is actually important. Since we're doing first order approximation I only care about the first significant digit and the number of zeroes, so for the suit:
Erest-iron = 20kg * (3*108)2 = 1018 joules
Enuclear binding-iron = 10-12 j * 300 mol * 6*1023 = 2*1014 joules
OK. Not really close enough to care about that extra energy. But is the difference between iron and nitrogen's binding energy enough to care about? This could be a bit tricky, since the total binding energy of the air depends on how many nitrogen atoms the suit has to absorb, which is dependant upon the difference in binding energies between iron and nitrogen, but let's just assume to start with that it's transmuting 20kg of nitrogen too (I'm making my approximation numbers more precise here).
Enuclear binding-iron = 1.408*10-12 j * 358 mol * 6*1023 = 3.024*1014
Enuclear binding-nitrogen = 1.233*10-12 j * 1427 mol * 6*1023 = 1.055*1015
Ediff = 7.526*10^14 joules
Putting that in context the difference in binding energy between 20kg of nitrogen and 20kg of iron is 0.07% of the total energy.
We need very slightly less air than straight rest energy conversion would suggest, but not really enough to be noticeable. We can assume that we need 20kg of nitrogen which translates to:
20 kg/1.25 kg/m3 = 16 m3
of nitrogen gas (as sphenning noted in his answer).
Now onto whether the suit can actually absorb that much air in any reasonable length of time. I'm assuming that the suit acts as a perfect pipe into a perfect vaccuum. While the size of the suit will change dynamically I'm just going to look at the two extreme cases of bracelet suit and full body suit. I'm doing this to simplify things, since my flow dynamics is weak.
The flow of air into the suit will almost instantly reach it's peak velocity (the speed of sound). It will also reach it's peak mass transfer velocity. At that point it is 'choked' and more air won't be forced into the bracelet any faster. We can calculate the volumetric flow, therefore, as
Area of suit * speed of sound
Assuming the bracelet represents 0.1 m2 of area this means we can in theory shift 34.3 m3 of air per second, or that our suit can be fully deployed using nothing but the bracelet in about half a second. If the full suit is 1.8m2 then it can get the requisite air (to build itself all over again) in 0.025 seconds.
Would this classify as wind?
It would also have some unintended side effects. Firstly: The noise would be immense. Secondly: Nearby objects would be thrown about (or potentially broken) by the shockwave of low pressure air moving outwards from the suit. Thirdly: once the suit is done deploying there will still be an awful lot of air moving in towards it, rushing to replace the removed air. This new air would then find itself rushing headlong not into a vacuum but instead into a solid super-suit. It would rebound and head back out as a second shockwave of air that would manifest as a very loud bang.
In essence: This suit wouldn't just cause wind. It would also cause a thunderclap.