alt Hacker NewsHuh, my initial expectation was wrong! I figured (even until close to the end of the article) that the problem was a dramatic increase in the amount of wi-fi or other 2.4GHz traffic in the area leading to interference, some of which was blocked by rain thus allowing more stable local connections.
I still think that's actually what happened here. A tree in the middle of a 2.4 or 5ghz beam would have an effect, but its branches moving by a few inches would not have this much of an effect.