I feel a more apt metaphor for the Sokal Hoax is a security researcher purposely submitting a backdoor to a random driver. If the commit gets accepted, the failure is not a result of code review being a flawed process or Linux itself being a failed project, but the driver maintainer who did not catch the vulnerability. Even then, mistakes happen.
One can also put some of the burden of responsibility on the security researcher herself, since code review/peer review is generally faulty when one acts in bad faith. Similarly, the Sokal Hoax is a great example of bad faith used to justify an ideological position through point scoring rather than legitimate criticism.