First off you would need to find yourself an Apple developer. Ask them what is possible. All you need is the ability to convert voice to either a UDP, TCP or HTTP-get message.
Bear in mind that you will still be using generic dictation based speech recognition that does not know anything about your commands or payloads so the accuracy will not be great, and you can expect failed recognition for many of the less common words or phrases (especially artist / song / albums etc.). This would be roughly equivalent to using auto-voice on Android, unless you are a believer in iOs and think they have better speech recognition than Android. I don't know or care which is better, but I doubt there is a big difference. In terms of recognition accuracy you are probably still better off using an app like Mobiola Headset (maybe there are other options now). This app has drawbacks in terms of UI convenience and battery consumption, but will probably work better with VC than Siri would.
If you are thinking that there is a way to create something equivalent to VC that runs on an iPhone, then it may be true but it would be a whole completely separate large-scale project and it is not one that I will be undertaking.
The final and best option would probably be an app similar to VoxWav made for iOs, but I don't have the tools or the skills to do that so unless another developer expresses an interest in making this, it is not an option.
