Then again, refraction can be quite convenient for FE'ers when it makes a round earth appear flat.
i get the feeling that "refraction" is one of those convenient, throw-a-way "explanations" that TPTB trot out to try and dodge "the hard balls";
although "refraction" is a relatively simple concept/explanation, TPTB and their innumerable shills are quite content to spew out a barrage of meaningless scientific gobbly-de-gook to baffle and confuse the average punter...
most people haven't got the qualifications, the time or the inclination to penetrate these "smoke screens";
that's the way TPTB like it......keep every-one distracted with mindless entertainments and too busy working to earn a living to have any spare time to consider the fact that they're trapped in "the Matrix";
(the word "amuse", for instance, actually means "don't think" or "anti-think"....'muse' being the Græco-Roman term for "thinking" or "pondering"/"reflecting on some-thing"....an "a" prefix is its opposite!)
as for refraction making a RE appear flat....well....where is the evidence for this?
where are the laboratory experiments (that have been double/triple checked and, then, peer-reviewed) to demonstrate this?
what are the refractive indices of air and water.....how do they compare and how can they distort an object's appearance.... 
most people know that water has a certain refractive index that can distort objects under-water......
does air have the same sort of effect?
quite frankly, i doubt it!
if any-thing, it would be only of a minimal, marginal kind that would make no appreciable difference
http://en.wikipedia.org/wiki/Atmospheric_refractionIn fact, not only does light refraction occur in laboratory settings, it is used by radio communications teams, from military applications, to commercial, to send high frequency radio signals greater distances than line of sight. The use of atmospheric refraction allows the signals to be sent at high angles as high as into the ionosphere, allowing the signal to refract back to the surface of the Earth beyond the horizon.
http://fas.org/spp/military/docops/afwa/U3.htmLight refraction is a very real phenomenon. So to determine the distance one can see from a horizon, one necessarily must use control, as you suggest right? Wouldn't that mean controlling the test to eliminate for any possible variables that could interfere with accurately measuring the distances seen, such as the refraction of light, or at least factoring in a margin of error due to it?
*****************
Also, as Mr. Bickles raises the point of repeated controlled testing and peer-review, wouldn't such mathematical models serving as proof have to match replicable and observable tests?
Pongo, can you please explain if your model accounts for why we cannot see the lower portion of mountains in the photos, or the bottom sections of towers, if the Earth were flat? Wouldn't a linear line of sight produce the entire image that we are looking at? Also, if your model is to be used to demonstrate that a spherical Earth were not possible, then wouldn't the sight seen from the top of the CN tower also have to be visible from the base of the tower? Or are you suggesting that atmospheric conditions prevent us from seeing visible light beyond a set distance from the surface? If so, then why doesn't this match observation? Light waves have varying frequencies, and atmospheric distortion would fade out the shorter wavelengths first. Thus, if this were to be what caused our vision to perceive a horizon, then objects should appear more red the further out they are, just as the Sun does due to the angle as it sets. Another method of falsifying this claim could be tested with radiowaves, as they can travel further than visible light. If we were to set up radio signals, and monitor them, would the distance of line of sight be further for them?
Please help me out in understanding your position on these.