Your "threat model" requiring the option shown to be least safe doesn't make its vulnerabilities less so. To develop a "threat model" one must first ascertain the system's intrinsic risk before leveraging one's specific needs. If the outcome is the least secure option, then only your risk increases, the security of each option remains unchanged.