I kind of wish to be wrong here, but doesn't that likely mean it also lends itself well to 'make any surface into a low-power, high-quality audio sensor'?
Any given electrical excitation can be either provided or detected by the part which couples the audio signal to the speaker aka microphone, and if this is an exception it is the first of which I am aware.
Of course we already have ok mics the size of sesame seeds so, what's your threat model?
> Of course we already have ok mics the size of sesame seeds so, what's your threat model?
Really hard to tell without a bit more specifics about implementation and adoption I guess, but co-option of ubiquitous consumer technology in the spirit of The Dark Knight (2008) comes to mind.
As you strongly hint at, there are ample opportunities for various kinds of unwanted listening already and it's up for debate whether this clever innovation brings anything qualitatively new to that. Large enough quantitative changes can often turn out to be qualitative changes though, and I find the present technology a bit suspect as something that might give certain people ideas about making shitty IoT devices with it and a partially different set of people ideas about getting huge amounts of those gadgets and sticking them in a lot of places where neither mic nor loudspeaker has any business being constantly present (and where currently there are indeed no such devices constantly present, because even people who aren't entirely in their right mind wouldn't pay for it all).
By the way, I am as saddened as anyone to see these aspersions cast on a fresh ingenious and exciting innovation. It ought to be that when we get a miniaturized, scalable, much cheaper and more power-efficient way to manipulate energy like this, that per default makes the world a better place. But here we are.