That risk doesn't make any sense when they can also just edit the page content on their end. Are we taking seriously the risk MIT would try and blame editing their own page on some kind of MITM that successfully masked true page content a) consistently b) for 100% of readers c) applied at an attack location in the network that someone had access to that isn't the same someone as controls the TLS-verified content hosting? This smells like rhinoceros-repellent levels of paranoia.
I'm not seriously suggesting this is likely, im just presenting a hypothetical scenario, and it doesn't need to be plausible in order for it to be used as a narrative. Do it properly, remove any doubt, secure your (one's) website so people _know_ the information came from you.
