Maybe they will someday release something with general availability and competitive pricing, but for now it seems to be behind a walled garden. But I wouldn't worry. If Wolfram Language truly exposes any novel concepts, then it is only a matter of time until the open source community implements those features for everyone else to use.
It's a shame, but I'm skeptical Wolfram will ever see the inherent value of getting real software into the world. From what I've seen, he would much rather build proprietary toys that only he can use, rather than building open-source tools that can be ubiquitous.
If Wolfram has cool toys and isn't willing to share, then so be it. The open source community can build our own cool toys.
imo, the language would be much more compelling if the syntax was redesigned from the ground up and made easier to debug (the error messages in Mathematica are next to useless)
And of course, I find value in other languages too.
[0]: http://c0de517e.blogspot.com/2013/10/wolframs-mathematica-10... (among a couple other posts about using it for specific tasks. He also spends some time defending it on there as well)