I think you are seriously abusing the long tail concept.
Chris Anderson defines: The Long Tail is about how our economy and culture is shifting from mass markets to million of niches.
(...) The article (and the forthcoming book) is about the effect of the technologies that have made it easier for consumers to find and buy niche products, thanks to the "infinite shelf-space effect"--the new distribution mechanisms, from digital downloading to peer-to-peer markets, that break through the bottlenecks of broadcast and traditional bricks and mortar retail.
The point that you are making in your post is almost exactly the opposite. You are saying that all these niche open source applications have no chance of survival if they don't manage to reach the head of the curve. You're applying the old economy pareto principle to open source projects.
The Long Tail principle applied to open source projects would sound more like: this myriad of small open source projects exist and can survive thanks to online collaboration tools such as SourceForge or FreshMeat that enable just a handful developers and users from anywhere around the globe to join forces around the same itch and scratch it efficiently no matter how tiny that itch is.
But Don't worry, we are all more or less confused but what the long tail entails ;-)