Statistically speaking ES has no value and way too much is made of it. Regardless of the number of rounds tested it represents only two shots, and in a statistically valid sample it should have almost no impact on the SD.
If the ES is low on a 5 round sample, then its possible to infer that that the loads are consistent but beyond that it means nothing, other than the ES values make up 40% of the mean and SD values. If the sample is 20 rounds the ES values make up only 10%, and have a much lower impact on the SD. In the real world the ES will increase as the number of rounds tested increases, but the SD may increase or decrease but to a much lessor extent.
To the OP's question, trying to compare ES between two different powders is somewhat akin to trying to compare the velocity between two powders based on 2 shot samples. What you see in the numbers is real, but what that means is very little to nothing about the real tendencies of the powders.
If the ES is low on a 5 round sample, then its possible to infer that that the loads are consistent but beyond that it means nothing, other than the ES values make up 40% of the mean and SD values. If the sample is 20 rounds the ES values make up only 10%, and have a much lower impact on the SD. In the real world the ES will increase as the number of rounds tested increases, but the SD may increase or decrease but to a much lessor extent.
To the OP's question, trying to compare ES between two different powders is somewhat akin to trying to compare the velocity between two powders based on 2 shot samples. What you see in the numbers is real, but what that means is very little to nothing about the real tendencies of the powders.