I reiterate my claim that it is a refraction that occurs, and not a reflection. Refer either to Wikipedia or to this page on atmospheric optics.
As for angle of reflection, as far as I know, it's always equal to angle of incidence.
-Erasmus
The angle of reflection is always the angle of incidence, but at a certain point the light hitting the medium won't be reflected, it will just pass through. This is why mirages will generally appear in the distance closer to the horizon; if you look closer to yourself, the angle of incidence will be high enough that the light won't be reflected (or refracted back up above the medium).
My understanding of mirages is and has always been as follows: Take a pane of glass and hold it in front of you, so you can see through it. Slowly tilt the pane of glass away from you until it is nearly level. At some point you will no longer be seeing
through the glass, but instead a reflection of what is above the glass.
This is what happens in a mirage, and the definition of wave behavior as being refraction instead of reflection is, in my mind, semantic.

It is quite clear that the mirage in this photograph is creating a reflected image of the car above. Whether the light bends (refraction) or bounces (reflection) when it hits that medium interface is, frankly, immaterial. You're right; all the academic sources I've found cite refraction as the cause of the image produced, but the effect is similar enough to reflection so as to be a non-issue.