That's very valuable information. Yes, it should have been heading to the center of the field after rounding the last corner. The KML file makes it look like it straightened out and was heading for the opposite end, but that's not what I recall seeing in person - it was, from my perspective, curving toward its run to the center.
I hit the gear switch miliseconds before it crashed, foolishly hoping I could get it down in time to shield the camera since at the time I thought it was completely out of control. I should have hit the stop button or the mode switch. Bad piloting on my part.
With all the input I've received here I'm heavily leaning toward it curving itself into the ground to round out the vertical turn for the center-field climb. The weather is going to keep me grounded for the next few days but as soon as I get a chance I'll hover-test it to be sure it's still airworthy, and then I'll do some steep vertical CCC work at higher altitudes to see if it dips down in a similar waypoint setup.
I really appreciate everyone's input. Thanks very much!