We've been doing some digging through this over the last week or two, and I think that the explanation is not so dramatic.
I believe that the simple time delay between real world events and Ohloh updates can explain almost all of this apparent decline.
When a project moves from one source control system to another, it often takes a few weeks before Ohloh is updated to the new location. This is a common event, so there is always some fraction of projects for which Ohloh's recent statistics are missing. With the rising popularity of Git and other distributed SCMs, this is especially pronounced lately.
Also, Ohloh sometimes simply has technical problems that prevent us from analyzing some repositories. It can take a few weeks to fix these problems. So again, there is always some fraction of projects for which Ohloh has no data over the last few weeks. Recently, a few large and prolific projects have been in this category.
These two facts alone can account for anything from a 10 to 20 percent drop in the apparent activity over the last two months. I would always take the most recent data from Ohloh with a grain of salt. Ohloh is eventually correct, but our global open source snapshot usually lags reality by several weeks.
All that said, I have a suspicion that the Ruby decline might be real. Time will tell.