Wadena,
See the attached image and tell us what we're missing. Apologies- I don't have good graphics software on this home computer. Still... I drew identical profile shapes below the waterline(s); the only thing that differs between the left and right profile is the height of the ends above the maximum possible waterline. The two horizontal dashed lines represent two possible waterlines. As you can see, wherever the waterline is placed (as long as it doesn't go above gunwale height), the in-water profile of the canoe remains unchanged. Therefore, end height can have no effect on performance with respect to the portion of the canoe that is in the water.
Fitz- those crooked canoes are something else, huh? I remember photos of Eastern Cree where you can plainly see crooked canoes alongside other canoes that have much, much less rocker. These people apparently produced multiple types of canoes, likely for uses under different conditions. The crooked canoe would be extraordinarily maneuverable, but it would surely spin like a top in high winds. By the way- by virtue of its "crooked" nature, it has bery high ends, but its performance within the water is not because of the high ends, but rather because of its extreme rocker. Here, high ends are simply a consequence of the rocker.
Personally, I like high ends. They add a fanciful measure of beauty that I like (in moderation- a Kennebeck Katahdin, for example, is a bit much for my taste), but I can understand how others may prefer the flatter profile of a guide-style canoe. In a canoe with high ends, I've certainly been blown about when paddling with little load, but the height of the ends in the air is reduced with added load, and a bit of heeling allows the wind to pass (that's actual wind, the breeze, for those who are giggling at "wind to pass") with less effect on the canoe.