> Thankfully, I've not felt the need to even consider this approach for the past 20 years.

That's because of all the branch predictors, probably. ;)