Reloading is a manufacturing process, be it on a smaller scale it remains a process that requires control.
To control processes, manufacturers record numerous data points / inputs which are diagnostic of that particular process and will be indicative of whether that process is in control or about to drift out of control requiring some form of intervention.
Most serious reloaders record raw materials just like any manufacturer does. Brass, powder batch, primers etc. An like with many manufacturers we are unable to test many of the components as we simply do not have the facilities and therefore we are reduced to trusting that that manufacturers "process control" is excellent. These manufacturers use statistical quality control as one of the means of "process control".
So many reloaders use some of these methods of quality control to achieve process control. Process control is about being able to control a process within a prescribed bandwidth, the tighter the bandwidth the more the cost and the time will increase to achieve these results. Process control is also about having the history of the process that allows one to immediately identify what has gone wrong and how to repair it. Many questions I see asked in reloading forums could have been answered by the poster had they taken the time to understand and practice the most basic of data collection and analysis. People spend loads of money groping in the dark and guessing on changes that they think need to be made in loads. It is also why people blow up actions with poorly designed loads etc.
Now if you are a hunter and you define your process as being 1MOA +- 0.5MOA as being an acceptable group then your process will require less input in terms of case internal volumes, powder dispensing etc. So simply put your cost will be less than people like me who like reloading to the nth degree. So set up you process control for your requirement. Quality is after all defined as "Fitness for Purpose".
Back to ES. ES measures the accuracy of your loading process and the quality of the components you use in this process. As Optimum Barrel Time is used to calculate accuracy nodes we should measure this ..... but we cannot as we do not have the equipment. What we can do however is to measure muzzle velocity which provides a contant measure and is cheaper to monitor and is a function of the OBT. So the further you deviate from the average muzzle velocity the less control you have over the process.
But ES alone is not enough so we assimilate all our readings to calculate the average velocity and then how many standard deviations we vary from this average.
To put std deviation it into perspective, world class manufacturers (all manner of goods) strive to produce to what they call 6 Sigma (or a total of 6 std deviations), which would relate to +- 3 std deviations about the average.
Now ES for me is the important number, the lower it is the better the control. Std deviation is merely the summary of all the results that make up the ES and is a function of ES rather than the other way around.
All my results (all load development is at 100m) show that the smaller the ES the better the group WHEN I am shooting properly. Recently I had std deviations on .375 loads of 1.0 / 1.5 and 5.1 std deviations but poor groups, later I realised it was poor left hand control which I could replicate. Firmed up my left hand grip and groups closed up, I had a 0.24MOA with a .375H&H.
A low ES, simply put, allows you to measure the accuracy of your loading process and removes a very important variable, the quality standard of the ammo. If my SD's are below 6 and I cannot get groups then the problem is me, proved it time and again.
A bit long winded, but at times I think we don't understand what a great tool statistical quality control is for the reloader.