There are some people who make a big deal about THD measurements throwing away "time domain" information because the instruments used to measure it perform signal averaging. This is often used in arguments regarding cables and supposed cable generated anomalies to bolster their point of view. It is erroneous and I'll explain why.
What is THD. Simply put, it is "Total Harmonic Distortion" - or distortion products that are directly related to the fundamental waveform. They are generated when the fundemental waveform departs from a perfect sinewave, which has no harmonic content. Therefore, by their nature, THD products are a steady-state.
If you are going to measure very small values of THD, you'll need to push the noise floor out of the way so you can make a more accurate measurement. So how do you do that? AGVERAGING! Noise is a random event, by averaging the signal over time, noise is minimized(rejected) while the desired information is retained. This yields a more accurate measurement of THD.
If someone wanted to measure a sporadic signal, this wouldn't be the way to do it, which makes me wonder: why anyone would bring up this argument in the first place?