In the end what counts is the result. Nobody cares what language you used if your software is great. Similarly, nobody cares about your language or the software, if it is a piece of crap.
Every time there is a post about D on Slashdot, Reddit, etc., there are people saying that D is such a wonderful language that they have migrated/are busy migrating all their projects to it. My answer to them: where is the result? Name at least one useful piece of software written in D?
On the other hand, the amount of software using a language and the decision to use a language is mutually reinforcing, and both are only loosely correlated with the quality of the language.
Ten years ago, there were almost no successful projects using Lua, but now there are many, and the language has barely changed at all -- likewise for Erlang. D is newer than either of these and occupies a niche in which other languages are deeply entrenched. I'd give it more time before writing it off.
Every time there is a post about D on Slashdot, Reddit, etc., there are people saying that D is such a wonderful language that they have migrated/are busy migrating all their projects to it. My answer to them: where is the result? Name at least one useful piece of software written in D?