What They Saw
This strain only has Mo and Fe nitrogenases, and they knocked out the latter. So when they removed all Mo from the medium and added W, the only thing it would produce would be Mo nitrogenase with W in it, theoretically.
Adding tungsten almost completely stopped acetylene reduction activity in the wild-type, unless 10x more Mo was added. So tungsten doesn't allow that kind of activity.
Mo also induces production of the Mo nitrogenase (which makes sense), but W also seems to have that effect, though not to the same extent.
Then they purified nitrogenase that had been produced when only W was present (in significant amounts) in the mutant. Purification was the same with this as with Mo in the wild-type, though of course the yield was lower. But the protein had 1 atom of W and none of Mo pretty much, as hoped. Still, it seems like there's only one FeW cofactor, not two, but at least there aren't any FeMos.
This protein still wasn't very active in acetylene reduction, and it didn't seem to be able to fix nitrogen at all. What it could do, though, was produce hydrogen when under an argon atmosphere, so electron flux wasn't totally abolished. And this activity wasn't inhibited by acetylene, unlike in the Mo version. Still, the amount of hydrogen it produced was only 1/4th the amount the normal enzyme could produce in the same conditions.
Finally they wanted to see if rhenium (Re) could take the place of Mo as a FeReco and be functional, since Re is not far from Mo and W (one more proton than W), so they tried growing cells with perrhenate (KReO4), but it didn't seem to help at all. The cells didn't even appear to be able to assimilate it, so it wasn't possible to test whether the nitrogenase could use it.
What This Means
It's difficult to do this kind of study, because Mo and other metals are almost impossible to eliminate completely from the medium. But they seem to have succeeded as much as possible, and still the nitrogenase only had one FeW cofactor rather than two. But surprisingly it showed some proton reduction activity, though not an exceptional amount. It seems like W in the protein is much more difficult to reduce, an essential step in the catalysis. Maybe the amount of reduction that's possible is only enough for some proton reduction activity.
One interesting speculation is based on an observation that Methanococcus thermolithotrophicus can fix nitrogen in the presence of tungstate at 60ºC, a pretty high temperature, so they wonder whether the nitrogenase with FeWco in R. capsulatus might also have more activity at higher temperature, but they didn't actually test this. Maybe another study.