Well, the final results (at least individual placement-wise) are now posted, and I am surprised by a couple things but not by the names near the top (maybe the order, but not the names). Zack's score is corrected from the initial posting, to make much more sense (since being 25 points off 365 means missing a 20 or not finishing a 20+5 or a 10+15 or something similar) but that also means I was that much better than the field in my pursuit of the threepeat. The wait to hear results probably helped contribute to my "need" to write some Hashiwakakuro so there are some good things that come out of having to wait. Now I can spend my tax stimulus check and celebrate somehow.
Statistics aren't up at the moment I am writing (they may be soon) but a quick search of recent USPC's shows the least completed puzzles on USPCs I took (2004+) was the atomic fusion at ~1.8% in 2006 with the Numerical Jigsaw (2005) at 2.5% just behind. Including the one "Google" USPC I did not take (2003) shows that the absolute low was indeed in 2003 with the Rotator Mosaic at .5% and Spellbound IV (.9%) and Dutch Week (1.2%) also amazingly low that same year.
Given "puzzle selection" possibilities in this age where a lot of sudoku-like puzzles will feel familiar, I wonder if Dot Triangles will set a new low?
Edit: Dot Triangles came in at 2.9% so Atomic Fusion still holds the win in USPC's I've competed in.