I suspect Solar Designer wrote the implementation for John the Ripper with the goal of cracking passwords, not validating them properly. For this purpose, slightly-truncated hashes should work just as well (maybe slightly better).

If Openwall and py-bcrypt are using JtR code for actually validating them, that's a questionable bit of software engineering. JtR may not be doing the same type of input validation that one would want in your authentication code. More evidence for this suspicion is that the input length disparity the blogger Rondam describes.