The only portions of the balloon that were retrieved were the lowest portion and the skirt. They came in 2 pieces, only connected by the string that attached them. The skirt had become completely unglued
It would be conceivable that the skirt came unglued prior to the burst, but it would complicate the explanation of the burst. My estimation would be that the rupture caused rapid contraction of the balloon, which caused the by then frozen glue to stop adhering.
The line of the balloon burst was surprisingly clean. It was nearly uniform, appx 6 inches above the adhesive line for the skirt, with one spike upwards to about 2 feet. It's possible the upwards spike was the weakened spot where the glue from the skirt had to be peeled off the skin. The uniformity of the tear was a surprise to me. It clearly indicated that there were no weak spots along the "ribs" and there was no shredding at all.
From what best I can tell, we have a burst primarily due to overinflation, although I wouldn't discount the effect of a weak spot determining the burst initiation point.