Encrypt word X to /etc/shadow encryption

It doesn't matter. The behavior you're seeing is implementation specific. That's why usermod does one thing and passwd does something else. They're different implementations. See what happens on Solaris, AIX, HP-UX, True64, Xenix...

The password field is a crypted or hashed string. When the user supplies a password it is crypted or hashed according to the algorithm specified in the password field. The authentication is only successful if both the supplied and stored crypted/hashed forms match.

A single ! or a double !! can never match any crypted password. In other words there is no input that will ever crypt to the result value ! or !!. Any string that could never be the hash result will "lock" the account. It may as well be foo or Mr. Spock.

Also note this comment under the --lock flag in passwd(1) on Linux:

Note that this does not disable the account. The user may still be able to login using another authentication token (e.g. an SSH key). To disable the account, administrators should use usermod --expiredate 1 (this set the account's expire date to Jan 2, 1970).

Setting a password with passwd or chpasswd generates a random salt, so users who happen to have the same password would not have identical hashes. In order to have identical hashes this way, you'd have to have a misconfigured system that somehow doesn't save entropy between reboots, and systems that are so completely identical as to repeat the random seed (that can happen with VMs, especially when resuming from snapshots), and for all the passwords to be generated after reading exactly the same number of bytes from /dev/urandom. This is highly unlikely.

Thus, if you see identical hashes, it means that the hashes were copied. The passwords were deliberately made the same, rather than by coincidentally setting the same password multiple times. Either the administrator directly edited the password database and copying the hashes, or they used something like chpasswd -e to supply hashes.

The only bad consequence of repeating a password hash is that it makes it apparent that the accounts have the same password. With distinct salts, as is normally the case, it would be impossible to tell that two accounts have the same password except by guessing the password.