I found this study to be amazing in a couple different ways. First of all, all the worst mutations seem to form spontaneously under selection pressure. I think we can predict the direction this virus is going, to include all of them, as the next step at least.  

I knew S477N was characteristic of Australia, but had not realized this was an especially transmissible variant.

To be honest, I did not like the way the article ends.  It is a scary idea, that masks are creating selection pressure and favoring mutations with higher binding affinity.  It's really just a guess.  In addition to this binding affinity hypothesis, it could also be that the main effect of all these mutations is to evade certain antibodies.  Then the virus can spread more easily, leading to higher viral loads which is what we actually measure.  The higher viral loads means people are shedding more and this is why transmisibility is higher. This is an alternative theory to the one in the paper, based on binding affinity.