Good study, but very skeptical of the "fleece is worse-than-useless" interpretation offered by @CNN. Here's the key figure from the paper. Indeed, fleece doesn't look *better* than control, but equally it doesn't look obviously *worse*. https://twitter.com/AmeshAA/status/1292559794407170055

Note that in theory this paper is a mere proof of concept (N of *one* in many cases, including the feared gaiter), and the authors say so in the paper. So far, so good. Good to have low-cost testing setups. But then not okay to promote your results to the press as a policy recs!

Also, as long as I'm being Reviewer #2, ratios (here, relative droplet counts) should be plotted on a log scale.

Because both measurements are noisy (note the large error bars on both) The apparent "difference" could easily be noise.