Oh, and while I'm at it: I'll take a stable C++ app over an unstable Mono app. I'll also take an unstable C++ app over an unstable Mono app.
You're just SAYING things without providing real world advantages over existing tech (that does the job better anyway).
Can you provide any real compelling reasons other than 'stability', which is really subject to how well the developer does his memory management? There is always a potential for said developer to do it better than a VM / Interpreter layer because then he can optimize for his specific use case.