Based on my quick and dirty benchmarking, it's similar in terms of performance. The vector comparison is consistent, the unpacking comparison is about 1.5x fast

Based on my quick and dirty benchmarking, it's similar in terms of performance. The vector comparison is consistent, the unpacking comparison is about 1.5x faster if it can exit early, but 4x slower otherwise. Here are the results: https://pastebin.com/QL4sNzhj https://pastebin.com/QL4sNzhj (the code is straight from the article, the benchmarking done with google benchmark).