This just provides the means for accurately measuring frequencies in the THz range.

This is important because previously it was possible to measure accurately either lower frequencies, until a few hundred GHz, or higher frequencies, from infrared to ultraviolet, i.e. from tens of THz to hundreds of THz.

Measuring accurately frequencies allows accurate spectroscopy in this frequency range, which can provide information about the chemical composition of materials.

'Frequency combs" are devices that can do at much higher frequencies what digital frequency dividers can do at low frequencies, and this frequency division function is what allows frequency measurement, by converting the high-frequency signals into low-frequency signals, whose frequency can be measured with the classic methods.

The method described in the article has the additional advantage of high sensitivity, i.e. it can measure very weak signals, which could not be distinguished from noise by less sensitive methods.

There are natural astronomical sources, from blackbody radiation. It doesn't penetrate the atomosphere so space based observations only.

TFA mentions spectroscopy and non-destructive scanning instead of X-rays.

Yes, what is really needed is a way to baseband a good swath of spectrum, e.g. in the neighborhood of 100MHz to 10GHz, so that conventional electronics can be used to study something more than simple low rate there/not-there activity. And conversely a way to modulate similar bandwidth onto arbitrary frequencies up in the THz region.

What's an antenna if not something that can receive radio signals?