Eh, the example of molecular velocities isn’t one of waveformes at all, casdave, which shouldn’t be too surprising as in and of itself, the concept of RMS value has nothing to do with waveforms per se, but rather with anything in which there are multiple measurements. In the examples you’ve been discussing, these multiple measurements are measurements at different times, while in the example of molecular velocities, the measurements are just different molecules. In fact, I’ve actually seen RMS crop up a lot more often in terms of discrete things (like molecular velocities, or errors in a regression line) than in terms of continuum things (like waveforms).