The man in the middle can do the key exchange with you. (You seem to be assuming that you start with a connection that you know to be reaching the desired endpoint and that you can start the key exchange there. You can't. The MITM intercepts the initial connection and does the key exchange with you and then turns around and initiates a secure connection with the eventual endpoint. The endpoint sees valid, encrypted traffic and you see valid, encrypted traffic, but the MITM attacker has the full plaintext communication.)

That's why you need the identity portion to be tied to the key exchange.

Even without the CA system, SSL (and TLS) is secure against _passive_ MITM attacks; That is, if the attacker cannot alter the data-stream. This is still reasonable for many simple links, but for internet, it is not.

If you've got no verification of the opposite party, literally any computer could be on the other end. The key exchange mechanism is secure, certainly, but any computer can execute it. The attacker can even show you the real website just proxied through, but that's plenty enough to get your password or any other access that the attacker wants.

A "passive MITM" attack isn't MITM any more - it's just interception. MITM means that the man in the middle is able to control the data stream that both ends see.