I'd expect significant amount of distortion to go along with the super refraction that would be required in the case of a surface level duct, but I don't see that here:
The mirror flash is in the red circle in the image above, and I do not see secondary visual evidence of enough of an inversion layer to produce a ducting effect.
Using Walter Bislin's model, there would have to have been a temperature inversion of around +0.05°F/ft temperature gradient at the surface in order to produce a curve-following "super refractive" layer.
I can't vouch for what was happening on that last afternoon on Monterey Bay, but we have been seeing a consistent, daily surface inversion in Southern California. It's what was responsible for this very interesting and cool sub-ducted "green flash". You can actually see the top of the inversion where the haze line is:
~photograph by Jim Grant, 2 Nov 2018It also caused a Fata Morgana; a distorted image of Catalina Island that I captured in photographs from San Diego at sunset a couple of weeks ago (75-80 miles away):
Here, you can see the layer of haze that is trapped by the inversion.
And I believe it's what is causing the power station structure in my Carlsbad viewing from La Jolla to look abnormally tall. Everything between the apparent mirage and that white line is, I believe, mirage (combined with towering/looming.)
But in my case, it's not producing a duct. Instead, I'm getting an opaque layer that isn't providing me extended visual range beyond a geometric horizon. As noted in that link above about ducting, an inversion layer won't always producing a ducting effect, and it's not in mine. But IF it's the explanation for why that mirror reflection is being seen, I think we should be able to see the secondary evidence of where the inversion layer is, like a demarcation of distortion at the top of the layer in the background power plant. I don't think such a layer would extend through to the top of the image. That would be extraordinarily deep for such an inversion.
The real test for flatness will be to see if that observation result is consistently achieved on different days under different weather conditions; especially clearer ones. The video author admitted visibility was poor.
The next opportunity I get, I'm going to try the same afternoon mirror method across a 13-mile over-ocean span between La Jolla and Encinitas, and the 20-mile span between La Jolla and Carlsbad.