Besides just micro-benchmarks, it’s obviously also interesting to see direct real-world impact of such changes. Not a lot of websites are dominated by for…in peak performance in a way that it makes a real difference, but there’s one interesting example that seems to stand out: The time spent in V8 during page load of meta.discourse.org seems to drop by like 13% with these changes:
That will be neat…
Obviously we are huge fans of Benedikt. 
Do read his other blog entry it is quite essential
http://benediktmeurer.de/2016/12/16/the-truth-about-traditional-javascript-benchmarks/
and the followup
http://benediktmeurer.de/2016/12/19/following-up-on-javascript-benchmarks/
TL;DR they spent too much time optimizing for benchmarks that didn’t actually represent real web performance! That’s why Google Octane is deprecated…
