I think someone needs to inject some reality into this discussion.
Firstly, this thing has zero application in a nuclear reactor. For starters, this alloy "works" by oscillating its temperature about some point, so that you get that magnetic switching effect. In other words, the alloy itself has to be cooled, since once you heat it up past the magnetic threshold, you need to cool it down past that threshold in order to heat it up again (unless I'm totally missing something here).
Upon rereading the article, again, it seems like you've got it spot on. Which makes it appear to me that essentially for all intents and purposes aside from very specific applications this is just a new way to make a thermocouple*, at least so far as electrical generation goes.
*Obviously it is quite different from a thermocouple in practice, but unless it is much more efficient I can't imagine a power generation situation where a thermocouple couldn't do the same job.
Secondly, you can't put a nickel alloy into a nuclear reactor - end of story. Nickel, compared to other metals, is extremely prone to neutron radiation damage. Statistically, a nickel material in a thermal fission reactor would experience at least one dislocation of every single nickel atom in the material every six months, due to neutron interactions (a dislocation is when the atom is ejected from its position in the crystal lattice of the metal). This leads to a bunch of material degradation phenomena, like creep (along all axis), large loss in ductility, large decrease in yield strength, large increase in the temperature dependency of yield strength, accelerated fatigue crack growth, etc. Nickel just isn't used in these cores.
Good to know, henceforth all my points about using this alloy in a nuclear reactor are conceded.