One might argue that if he's specifically following the direct instructions of the ATC, who is a licensed pilot in this case, that the ATC is "the pilot," and his allowing of a non-pilot to push the buttons on his behalf was a necessary deviation to handle the emergency and land the plane.
I’m far more familiar with operation under Part 107 (sUAS, i.e., drones <50#), but this seems consistent with how the FAA has it structured.
I let kids fly my quadcopters regularly. Legally, I’m the “RPIC” - “remote pilot in command”. The fact that someone else is physical operating the controls is irrelevant. I’m responsible for the safe operation of the aircraft. Because it’s remote, there is a requirement that I be able to take immediate control of the aircraft if necessary; standing next to the person with the controller is sufficient to meet that requirement.
That's an interesting twist -- I really hope nobody at the FAA gets such a wild hair to try to pin a certificate action on the ATC controller (here or ever, in any situation), but --
There is a SODA precedent for what you describe. As I understand it, a deaf pilot can get an instrument rating if they have a translator onboard to run comms for them and translate to ASL in-cockpit. So this is sort of the same situation you're contemplating, with a different chair position -- "translator" in the ATC room instead of the right seat.
I do feel the regulations allow all sorts of circuitous logic, that any "outcome" could be achieved with enough incentive. ATC is the last place on earth where I feel cooperation is the inviolate rule of the day, so I'd hate for some maverick prosecutor to throw ice water on the pilot-controller relationship in a witch-hunt for blame.
there's absolutely no way the controller qualifies as PIC for this event - to be PIC of a non-unmanned aircraft you would have to be onboard the aircraft.