Thank you guys!
The METAR data is pulled from NOAA - they expose current (cached) data for the whole world in CSV and XML (AFAIR) - I'm pulling one of these files. It's stated in the about section. The service is generally written in such a way where adding new data sources (be it APIs, random data files or whatever) is modular and relatively easy.
Regarding the domain I was surprised too. I felt that ".org" is more suitable to a non-profit hobby project than ".com" (which theoretically is for companies). Maybe I'll buy the ".com" and redirect to ".org" if the site gains some popularity.