This is exactly Searle’s point. Whatever the room is doing, it is not the same as what humans do.

He fails to show that. All he has shown that the human+room-system is something different than just the human by itself. Well, doh, nobody ever assumed otherwise. Running a NES emulator on my modern x86-64 CPU is something different from running an original NES too. That doesn't mean that the emulator is more or less capable than the real NES or that the underlying rules driving the emulator are different from the real thing. You have to actually test the systems and find ways in which they differ. Searle's experiments utterly fails here.