Where is this 60% approximation coming from? Refraction only accounts for a few extra minutes of sunlight during the equinox, and the phenomenon is both widely documented. widely understood, and easily predicted - kind of disappointing magic. Somehow I don't think the real coverage is anywhere close to 60%. Of course you can't approximate it from what's covered by the Mercator projection maps without first correcting the projection, but I'm sure everyone already knows that.
While we're on the topic of the sun I have some other questions, which I'm sure have been asked before plenty of times, but hopefully less so than the the OP's:
1) If you look at a timelapse video the sun appears to move through the sky at a close to constant velocity. FET claims that the sunset is due a perspective effect of the earth moving away from the viewer. The way perspective works, the object's position towards the vanishing point is inversely proportional to its distance - for instance, railroad tracks appear to get closer together the further they are towards the horizon. In order for the sun to appear to move at a constant rate it would have to be accelerating away from you at a constant speed. And yet if this were the case it wouldn't appear to rise and set at a constant rate for all viewers. So how do you explain this?
2) According to Rowbotham, objects far away from you will appear to recede behind the horizon instead of converge towards the horizon. Indeed the sun can be seen partway intersecting the horizon, so this must be an example of this phenomenon. And yet, Rowbotham also says that if you look at the object with more optical magnification it will appear further away from the horizon. So forget boats and the argument that waves aren't calm enough for this - why can't this phenomenon be showed with the sun? Or is it just that none of you have performed this experiment yet? You can buy anti-glare filters for only a couple dozen dollars you know...