Is this just biased towards the type of projects done on Newrelic?
There are two data sets, one from newrelic and one from gemstats.org. So, in short, no, it's not. It is, however, biased to the Rails community and web-apps in general, but this was mentioned. If you know a way to get a sampling of non-Rails usage, feel free to suggest.