The key point here is integration though. Just putting a speech recognition app on a phone isn't doing much. But making sure the most common cases you are going to be using it for, e.g. appointments/reminders, searching for specific things, finding a nearby hotel etc. work reliably no matter in what way you phrase it, is the part that makes it really useful. Also consider how you can make sure this works as expected by using all the context available onm a modern smartphone (location/movement, weather, calendar, emergency information, congestion) If Apple does the polishing work for that to work it will be a huge success. They probably aren't there yet though, otherwise their presentation would have put much more emphasis on it.

I tested the current state of voice recognition on Android and was mostly disappointed (bad recognition, and nowhere the "intelligence" you would expect).