Tim, you posted while I was writing this so I made minor adjustments and added info for others.
Tim L, I am glad you posted these groups because they confirm a few things about the macro trend of the barrel if it’s a positive compensator.
If you are jamming, forget the free fight changes. That was for an off the land process. Run with what you have got. I often used the jam to induce barrel lift as I have said before. You may get a different compensation profile if you introduce free flight.
As Brad said use a tuner. A light variable one would suit your barrel.
On first observation your best groups have an extreme spread roughly 3 times greater compared to other groups extreme spreads, and in one case 4 times the extreme spread. For example, either side of string 2 (with an extreme spread of 24 fps), and is the tightest group in this section, strings 1 and 3 have extreme spreads of 6 fps. Either side of your tightest group overall, string 5 (with an extreme spread of 19) you have 12 and 13 fps extreme spread. Sort of contradicts the theory of loading for minimum extreme spread.
However either side of string 8, which is a peak node thereabouts (its between 7 and

with an extreme spread of 20 you have 6 and & 7 fps extreme spread. But string 7 is not on a node and has vertical whereas string nine is on a trough node and it is tight.
It is easy to conclude that extreme spread is not a big a factor in determining group size. The group size depends on its relationship to the node.
I have also said in previous discussion, that the trough node can produce extremely tight groups and your groups indicate that and are pretty rhythmic in their application and demonstrate this too. When using this node you have to be convinced you can deliver within the parameters of this load by your reloading technique and hope your environmental window will not tip you out.
In your case the lower node (String 9), the left hand group happens to be closer to an upper node peak, and the right hand group wants to introduce lateral. Between a rock and a hard place?
The safer bet, while not as tight as the trough node, is the peak node between String 7 and String 8, because you will not have the lateral in your case and so you will not be mislead by a perceived wind change as the day warms up. You may think you will get some vertical as slow shots are tossed higher if you creep towards String 7. But in reality, because your group has a positive compensation aspect, the short range elevation you perceive in testing will be minimal at the long target distance. I think a load of 47.3 grains will be sufficient to put you on the left hand side of the node and still group well and put you on the node as the day warms up and be back far enough to prevent you slipping in the negative direction.
A further clue can be seen in your spreadsheet analysis where you graph velocity against string number. You have a flat section between strings 7 and 8 where velocities don’t change much. Now all you need to do is identify the nodal peak, which you can do by laying out your groups side by side as I have. Then draw the channel about the major peaks and troughs. The channel shows the macro picture or the compensation profile of the barrel. Next, draw the dashed line following the sine wave of peaks and troughs for the micro picture. This will show you the best area to load when you consider the node and group patterns either side.
If you look up the weather conditions on the BOM site, and you reckon they are stable, be a risk taker and use the lower node, you might be pleasantly surprised. David.