Paracelsus

OK, so I am merely in the planning stages of my CarPC and still trying to cover all the basics in regard to hardware and software.

Trying to make sure I can really implement all the features I want and replace those I already have in the car with the new equipment before I shell out the money.

So, I'd really like voice recognition for a small, select set of commands. From reading here I think that voice control is still a bit dodgy in some of the frontends, or am I mistaken there? I know there is a problem with "background noise" and music replay.

Currently my stock car system has a steering wheel button to activate voice control, it mutes the sound and then waits a few seconds for a command. I am planning to wire in all my buttons via fusion brain or something, Would I be able to setup the common frontends to act just like that - mute sound on button press, activate "listening" for command and execute it and resume music playing? Or is the "configuration depth" of those system not as deep?

What would be the most open, most configurable frontend? Currently I am considering road runner, flux centrafuse and frodoplayer. They seem to be the ones that attract most people, and I hope that is because of stability and feature wealth.

Also, can phone implementation be considered matured or is it still a beta thing that requires a lot of setup and still would cause occasional hickups? What is the FE with the best phone implementation? (We are talking about Bluetooth connected Nokia here).