Thanks for highlighting this study...I've just quickly read the original published article.

The authors make a fairly compelling case that there are two circulating strains from their limited set of 100 samples, L and S -- defined by two "linked" mutations. The S strain is ancestral, i.e. is more likely to be the 'original' strain of the virus, since it's more similar to other similar viruses. Intriguingly, the vast majority of Wuhan isolates, sequenced in early January are L strain, i.e. the "new" strain. A significant minority of isolates from outside Wuhan are S strain. However, although the authors make predictions about relative severity, there's no actual data to support that...all we can say is that there are two strains, and the ancestral one is actually more common outside Wuhan.

However, this would be an excellent place for virology labs working with SARS-CoV2 to test the hypothesis...of whether one or the other is either more virulent or transmissible etc.

The authors don't make an unreasonable conjecture: that with more stringent control measures, there is selection pressure for a strain that causes milder symptoms...but that's not actually derived from data. And the L strain is still the more prevalent (70% of samples) overall, even outside Wuhan.

tb