First of all, I want to say that I appreciate the fact that Dogplatter actually has the balls and integrity to come back and continue to debate a post, unlike our dear friend Tom Bishop. Dogplatter, I still disagree with you, and I will show you why in just a moment, but I do think it's a nice change for an FEer to come back and continue a debate.
Now for my rebuttal.
I've put together a simple diagram which will help aid me in my argument.
Its purpose is to sequentially illustrate what an object would look like moving farther away from an observer in two scenarios - on a globe, and indeed as it does in reality in reality (Fig 1) or on a flat plane (Fig 2). The dark blue represents the sea as you may have guessed. With the light blue line, I am accounting for your "wall of water", purely for the sake of argument. Again, I believe there is an error here, and previous threads have dealt with it. But for the sake of argument, let's say that it is there. Let's say that there is this compound effect that rises slightly above and obscures the true horizon. Now this ultimatelly changes nothing, as you will see. Before I move on, I would like to point out that in both scenarios, the object has reached the distance at which it would begin to 'sink' on the RE model.
On a globe, the effect you see (be it with a 'sinking' ship or a setting sun or whatever) is the one depicted in Figure 1. Once the object reaches a certain point (i.e. the horizon) it begins to 'sink' suddenly and distinctly and the effect does not last long. However, on a flat plane this would NOT be the case. On a flat plane, the object would continue to get smaller and smaller. Yes, given this 'wall of water' it would eventually be obscured, but it would be obscured by "shrinking" into the waves, not "sinking" into them. The same would be true for a sunset. I really don't know if I can be any clearer about this. It's really not rocket science.