I don't think we will ever turn wind on by default. It would provide a more realistic experience out of the box, but we would be inundated by complaints about how the models are acting all funny.
We could also delay sounds to simulate for the distance they have to travel from the source to the listener, for example. That would be more realistic, but people would only perceive it as a bug. There are a some situations where greater realism is actually undesirable.