Simply put, there are too many observations that go beyond the curve for the curve to be the reason we don't see things on certain days. And a really easy way to use logic against the curve is that long distance photography is usually done when the weather is favorable and refraction is as low as possible. For things to loom up over the curve, you first need to demonstrate light bending down (but that's a separate issue), then you would need a large amount of refraction for to do what people say it's doing to come up over a ludicrous amount of curvature.
