Dave Tosh points out that not only does Elgg still exist, 1.0 came out recently, it's winning awards, and doing quite well. Indeed, the pattern of making an initial splash, running into problems, retrenching and coming back a few years later is fairly common in software development (see also The Mozilla Project). Certainly SchoolTool has gone through a similar progression in a similar timeframe (although we're still behind them), and I assume at this point most people figure we've irretrievably run off the rails as well. Since I don't think anything other than a few good releases by us will change that impression, I don't spend a lot of time worrying about changing it in the immediate term.
Anyhow, in the case of Elgg, my original comment, that "their model went off the rails almost immediately" was not completely unsubstantiated. As I recall, Elgg was started around 2004 explicitly as open source social software for learning and schools (perhaps the former more than the latter). This was right when the momentum for educational blogging was getting rolling and many of us hoped Elgg would provide the right platform for the job. Before too long, there was a lot of noise from the Elgg team that open source for education was not working out for them as a business model, people I know who use the software complained about a lapse in support and development, developers seemed to be dropping out of sight, and ultimately Elgg was, I gather, retooled and remarketed as a more general purpose social networking application.
Which is fine by me, as long as it is still open source, and, in particular, people like Jim Klein can hack on it to do what they need at their schools. It is exactly the kind of collaboration I want to see.