>>23Turing completeness isn't enough for a language to be useful for game development. Brainf*ck is Turing complete, but you wouldn't want to use it for game development. It basically has no ability to interface with 2D or 3D graphics library (unless you create some custom solution where the library interprets byte-stream output at runtime), and even making text-based games like Rogue or Super Star Trek would be a nightmare.
And Lisp isn't what people think of as a gamedev language either (though it's far better than brainf*ck). Because it has a REPL, people assume it's exclusively an interpreted language, and therefore slow. And while many implementations do feature the ability to compile into machine code, it's too abstract to match the speed of asm or C/C++ (which have traditionally been the only suitable options for real-time game engines).