From the article:

Of course, there’s another possibility that takes us far beyond astronomy when it comes to making use of this important length: creating and measuring enough spin-aligned hydrogen atoms in the lab to detect this spin-flip transition directly, in a controlled fashion. The transition takes about ~10 million years to “flip” on average, which means we’d need around a quadrillion (1015) prepared atoms, kept still and cooled to cryogenic temperatures, to measure not only the emission line, but the width of it. If there are phenomena that cause an intrinsic line-broadening, such as a primordial gravitational wave signal, such an experiment would, quite remarkably, be able to uncover its existence and magnitude.

Isn't that basically an H-maser? Not something found every day on eBay, but not really all that exotic either. Every VLBI site has one or more.

Given a suitable state selection mechanism, which is what masers rely on, I don't see why it would be necessary to flip the states "manually" through ionization or any other mechanism. Keeping the state-selected atoms away from the container walls is the real trick.

I’m reading that an H-maser emits 1.4GHz. Maybe you mean something besides it’s emission frequency?