What classes of vulnerabilities are you referring to that are likely to show up in a microkernel-based design but not in a monolithic one? Do you think adding those classes of vulnerabilities to the picture will result in more, about the same, or fewer critical security issues between the systems overall, and why? Please be specific.
