I worked in the business of engineering such things. The factory for the AGM-86 was in the next building over from my office. I worked with ex-military experts all the time on subjects like this. You may view horseshit on Wikipedia, but I am relying on my experience and professional training.

It depends on what you want to call a cruise missile, but self-navigating long-distance missiles need to avoid ground clutter and typically travel at least tens of feet above the ground, if not hundreds. Yes, they can close on the target when they are close enough not to risk a collision with the ground clutter. Usually in a dive in order to suppress the effect of altitude error on the projection of the ground error. Vertical dive is best, but you can't go through a window without some slant angle (I've seen that done).

A battlefield missile like a Javelin does not fit the definition of a cruise missile.

But the size of the object in the video is not compatible with cruise missile dimensions. Nor was the resulting wreckage. Nor was the fact that clipping the light standard would have been fatal for the missile (which is why they don't fly that low). You want to focus on the most slender evidence and disregard all the other evidence and disconfirming circumstances. Sorry. That doesn't much solve real problems.

Now, if you can recount examples of site doctoring to accomplish a cover-up, that would be of interest. All I can say is that the 1976 mid-air breakup of the Canadian CF-101 resulted in a crash that was not covered up. Visible as hell. International incident.