I think having both would be the best option.
I'd argue that the necessary tech has been around for a while now and they probably concluded it's not a good fit.
Every drive-thru that has the monitor where you get to check your order for accuracy eventually starts displaying ads instead -- lowering accuracy and offering something you don't want.
It doesn't have someone whisper/shout through the speaker because they don't know how loud they need to speak to be heard outside. It speaks clearly and can be understood which is not always the case with the bored and disinterested high school workers the store tends to hire. And my personal favorite is that I'm not greeted with "McDonalds! Thank you! Order please!" which seems to be the standard greeting they're training employees with at that location.
I've only had it get an order wrong once (though the food does still get made incorrectly about 10% of the time) at which point it quickly switched over to an actual person to complete the order. You can tell it's more efficient too because at busier times, the non-human order taker lane moves cars through at a two or even three to one ratio.
Overall, I'm surprised that more locations haven't switched to the software already and especially that other drive thru businesses haven't begun testing the software out here.