Pretty much everything has to be just so with loaded .223 Rem rounds using heavy bullets in order to keep the ES/SD low. Regardless of what some may think, without a load that has a reasonable ES value, shooting a .223 with heavies at 1000 yd will probably be a long and painful day. It IS important. Precise charge weights, uniform neck tension, uniform internal case volume...all of these things are critical in order to keep the ES low with the relatively small .223 Rem case and heavy bullets.
In my hands, ES values for 5- to 10-shot groups for .223 Rem loads with Berger 90 VLDs are typically in the upper teens to around 25 fps. A bad run might be as high as the low to mid 30s, an exceptionally good one between 10 to 15 fps ES. Per Wade's observation above, I'd be very shocked to see anyone loading .223 Rem ammo with 10-shot ES values consistently below 10 fps, as can be done with .308 loads. Further, .224" bullets with relatively long bearing surfaces, such as the 95 SMK, can sometimes make it even more difficult for which to obtain solid ES/SD values.
There is probably not any one thing that will be the "instant fix" for poor ES/SD with .223 Rem and heavy bullets; meticulous brass prep and reloading practices are essential as I mentioned above. However, one thing you might try if you haven't already is sorting your cases by weight/volume. I'm sure there will be a few readers that see this and want to start yet another "case weight versus case volume argument". If so, please take it elsewhere and start a separate thread; this thread is for the OP to get feedback on loading .223 Rem. Nonetheless, I typically sort cases by weight when loading for an F-TR match. I also sort bullets by weight to remove any gross "outliers". I do not typically do either of these steps during load development or practice, only for matches. I figure if I can get things working the way I want without those steps, they will most likely only make rounds loaded specifically for matches even more consistent when added to the reloading process.
In the process of determining case water volume and case weight over the course of several years, I have become convinced that in fact, case volume is generally proportional to case weight. Notice I stated, "generally" proportional. That is to say, when you plot case weight versus case volume, you should see a general linear trend to the scatter plot data, of which the best straight line (regression) through the data points should have a negative slope. In other words, as case weight increases, case volume decreases. The relationship is not perfectly linear - i.e. there will always be a few outliers, meaning case weight is not useful for estimating case volume on an individual case basis. What it does mean is that by sorting many cases into specific weight groups (i.e. "light", "med", "heavy"), you will end up with more uniform internal volume within a specific weight group than if you had done nothing at all.
My main reason for even mentioning this is that I recently started a new batch of Lapua .223 Rem brass. I noticed with both 95 SMKs and 90 VLDs that my ES/SD values were larger than I typically expect, so I decided to weigh a few fired cases as I had not yet prepped any of this brass for a match (i.e. no weight sorting of it as yet). To keep things simple, I selected 5 cases from a single shot string that had obvious velocity deviation. The load was Berger 90 VLDs over Varget. Below are the velocity data and weights of the individual cases.
Velocity(fps)
2858
2853
2821
2823
2825
ES = 37, SD = 18.0, Avg = 2836 fps
Case Weight (g)
6.3010
6.3121
6.2238
6.1763
6.2274
As you can see, the two high velocity values that wreck the ES/SD are approximately 30 fps faster than the other three. Although this is only a sample of n = 5 cases, I have done enough of this in the past to know that those two cases will have much smaller internal volume without even actually measuring the water volume. In other words, this is no coincidence. As it turns out, cases that weigh over 6.3000 g in this particular Lot# of brass constituent about 25% of the total number of cases, a relatively high proportion. I am in the process of more precisely separating out and characterizing these heavier cases with respect to internal volume and their effect on velocity. Nonetheless, this is a good example of how case weight (and likely internal volume) can contribute to excessive ES/SD in .223 Rem loads with heavy bullets.
Whether you use case weight, or the more rigorous (but time-consuming) method of determining actual water volume, you might find that having more uniform case volume may buy you a little improvement in your ES/SD values. It's pretty easy to sort a few cases and see whether it helps.