Frankly, the idea that no app or company is doing this or selling the data would require an enormous conspiracy of secret suppression. The argument that one can't easily hash compressed sound fragments in a way that matches for copyrighted music or generic advertising terms and send that data back linked to a location/ID is sort of laughable.
A team of CS engineers at NSA.
1-2 people in privileged positions either at the point of assembly/programming deployment to ensure the code is delivered. - for each company which is basically just alphabet, Samsung, and Apple or their chip providers which is about five companies.
This does not seem all that impractical - how many people are dumping firmware from in the wild consumer cellular devices and reverse engineering enough of the code to see if this functionality exists or not? Anyone with that skill is likely making a killing doing other work.