The problem with all of these devices is that there is a lot of noise power in the world. A jet engine is loud because the acoustic energy it produces displaces large volumes of air very quickly. To cancel that energy, you also have to move huge volumes of air. Headphones are good at this, in part, because they attenuate so much energy in the first place with an over ear headphone, or well sealing IEM, and place the speaker very close to your ear. A source of acoustic energy which is D meters from your ear has to produce at least D^3 more joules than a speaker right next to your ear, to cancel the same amount of noise. It’s possible, kind of, with very big speakers, but a handheld device just won’t cut it.
And they inevitably turn every service into an ad delivery service. Thus, enshitification. Subscription services directly convert the utility of the application into revenue. It will always make a better platform, but it might not scale as well.