I'm thinking official D&D "virtual table" products qualify as DNF type candidates; notable for being in production at least 2-3 times, having been switched through 2 or even 3 editions now (maybe four if you count 3.0 and 3.5 as separate editions), the scope of the project has been repeatedly widened and cut back, no release in 6 years with a feature limited release being the most recent (and having taken 2 years even then), announced in a fashion that would indicate upcoming release followed by "well... it's not ready yet, so wait another few months", and reasonably wide anticipation. The best we've gotten recently on this front is Gleemax. And we all know how that's turned out.
Anyway, I wonder if this topic shouldn't be moved to General RPG discussion.
Wulf: If it's written, and in reasonably publishable form, but the market simply doesn't favor spending the work required to finish publishing it (like say a piece that was announced prior to the "no half-orcs in 4e" announcement focusing on half-orcs in 4e) it's not exactly in the same class of vaporware as DNF. That's not development heck. That's just bad timing.
True vaporware-osity comes from being something that if finished, would be feasible to sell immediately were it finished, meaning the only thing that stops it is either unrealistic claims (aka BS technical claims and other such stuff) or unrealistic goalsetting and poor decisionmaking by developers. The prototypical example of the first class of vaporware is a perpetual-motion machine. It is inherently unrealistic. The second is Duke Nukem Forever because the apparent intentions and expectations of the designers have been unrealistically high, and because the devs go back and scrapping things unnecessarily.