Hey OP, why didn't you just ask which is better, 9mm or 45ACP? lol. I have spent a lot of time trying both methods, the "Satterlee" method and OCW. Satterlee is nice because you just go shoot over a chronograph, you don't have to worry about a range, shooter error, or other externalities effecting the final load you shoot. However, while I think most things in life can have a mathematical doppelganger, there's usually more that can be learned outside of hard and fast math. So, while you may find a load with super low SDs, it doesn't necessarily correspond to a solid load that will stand up to differential environmental factors, component lot variances, etc. Take benchrest guys, they were producing disgustingly small groups long before chronographs so a good SD wasn't what they were after. The two are not always different, but they're not always the same either.
Second, I talked to Newberry about his method once and I don't recall him ever mentioning SDs, certainly not as the prime factor in load choice. That is not the metric by which all his loads are judged. However, the Ko2M guys live and breath SDs in the 2 to 3 range, so they do. Thus, I thought to myself, there must be a very real need for a low SD and mathematically I understand why. I just didn't see good SDs translate to good groups all the time. Now, if SDs are not the big dog in the pen (I've already pissed a lot of dudes off with that), what is? My answer is that both group size and SD (and ES) is important. My go-to procedure now is to shoot an OCW test and then choose the loads with the lowest SD, or thereabouts. Complex, right?
Recently I did this for a 6 CM load. Using 110 ATips and H4450, I made loads from 40.8 to 42.3 in 0.3gr increments (N=6 loads), then identified what I thought was the OCW. I tallied the SD and used that to solidify my decision. That's it, I didn't even need to fiddle with seating depth. In the end, I had what promised to be a good load so I loaded up 100 rounds in two reloading sessions and went for some long range "testing." After a hundred rounds, I had a load that would was in the 1/2 minute range (pretty good for me), an average velocity of 3032, and SD of 7.4. Time will tell if over numerous loading sessions the numbers stick, but a 100 shot sample size over two sessions is certainly more than most guys do.
Based off my own empirical evidence (which is subject to change as further data is collected on the topic), you'll find that powder burn doesn't necessarily produce linear velocity increases, hence a "node." But good groups don't always reside in the flat spots (there goes the rest of the guys). I think there has to be a happy medium. Keep in mind that seating depth, brass prep, component performance in a given barrel and chamber, and other factors also produce their own affects, so the conversation probably needs to be a bit broader than just powder burn. But, FWIW, here ya go.
