Agreed - not looking to use ES/SD to predict a load as the most accurate, that it's likely to be the most repeatable. An accurate load that's not repeatable from string to string is what I'd be looking to avoid, using SD/ED as a measuring stick for that likelihood.
Bud in a perfect world lowest ES/SD would win every damn time if all conditions were perfect, equipment even steven and there were no human interactions. It isnt a perfect world.
Good read here
https://www.angelfire.com/ma3/max357/houston.htmlExcerpt
For the most part, shooters arrived at the warehouse with troubles. Their rifles were inconsistent — one group in the teens, the next in the .3’s — for reasons they could not fathom. Others had consistent .25" to .30- something rifles, an accuracy level guaranteed to put a competition shooter down near the bottom of the pack. With the list of potential problems significantly narrowed by the elimination of moving air and dancing heat waves, the answers were easier to isolate in the warehouse, and shooters drove hundreds of miles or flew into Houston to get to the source of their tribulations.
ES/SD had nothing to do with it.