Are we sure AlphaZero has better learning efficiency than human?

Sure, it reached peak skill after 4 hours of learning, but how many games did it play during those 4 hours? How many moves did it memorize perfectly and analyzed? Are those numbers even achievable by a human in one's lifetime?

Even with AlphaZero's efficiency, it still evaluates 80000 moves per second, which is by far more moves than a human grandmaster evaluates in an entire game. If we cut AlphaZero's "processing power" to that of a human, can it still beat a top level human player, let alone other AIs?

To me it seems like there is still a long way to go to improve in this space.