Mac,
I've spent quite a bit of time running numbers and thinking about your argument for equal tof G1 BC's. To sum up my conclusion in words, I can say the following;
The fact that an 'equal tof' G1 BC can match the G7 based trajectory more accurately than the published G1 BC is a case of two wrongs making a right.
The fundamental fact that the G1 standard drag curve is so different from the bullet in question means that when the raw data is reduced correctly, the resulting predictions should be inaccurate (first of two wrongs). Finding a way to make the G1 and G7 based trajectories match by 'tweaking' the G1 BC to match the G7 tof is the second wrong which is required to correct for the first one.
Furthermore, don't forget that the range (or lower velocity, or tof) at which you choose to identify the equal tof G1 BC is completely arbitrary (just like the windows I defined). You can come up with an equal tof G1 BC that is 'optimal' for a given window, but has more error than the published G1 BC at ranges outside that window.
To investigate the matter further, I ran the following comparisons. Using my published averaged and stepped BC's, the following trajectory metrics were produced:
*Berger 6mm 115 grain VLD, 3000 fps MV, standard sea level ICAO atmosphere, 1.5" sight height, 1000 yards.
G7 average (drop/tof)
275.5"/1.408s
Stepped G7 BC (drop/tof)
275.7"/1.410s
G1 average (drop/tof)
279.4"/1.418s
Stepped G1 BC (drop/tof)
277.8"/1.419s
Of the above 4 predictions, I consider the stepped G7 to be the most accurate representation of the bullets' true trajectory. Second most accurate would be the average G7 BC trajectory (only -0.2" error). Then the stepped G1 BC at +2.1" error. Finally, the average G1 BC is the least accurate at +3.3" error.
What strikes me about the above comparison is the inconsistent effect that tof has on drop. For example, the difference in tof between the stepped G7 and the average G7 is 0.002s, and the difference in drop is only 0.2". However, the difference between the stepped and averaged G1 is only 0.001s in tof, but the difference in drop is 1.6". In other words, the tof and drop aren't correlated as one might expect when you compare trajectories based on averaged vs stepped BC's.
The inconsistency above can be explained in the following way: Although the tof may be the same from point a to point b, it is possible for the drop to be different based on the shape of the drag curve. And this is the best way I can describe why the equal tof method you're advocating is not a better way to reference BC's to standard curves when the shape of the curve is known. That last part is an important distinction. If you don't have any information about the shape of the curve (as in the case when you only measure overall tof), then you only have one single tof, no knowledge regarding the shape of the curve, and all you can do is derive a BC based on the single tof. In that case, I agree with Ken, it's the best you can do. However, I specifically designed my test procedure to measure tof in several increments so that I would have information on the shape of the actual bullets drag curve. This allows for the comparison/averaging practice that you say produces less accurate (predictive) BC's. In fact, the BC's that are defined with knowledge of the drag curve are more accurate than BC's that are defined without knowledge of the drag curve.
Thinking back over this debate, it's very easy to understand why you would believe so strongly in equal tof BC's. The fact that they have less error when compared to more appropriately referenced (G7) BC's seems to be compelling evidence that it's a better way, and if the original G7 BC's were defined based on a single tof alone that they would also be improved. But for reasons I described above, it's a case of two wrongs making something closer to right, for a specific case. If all you care about is the result at a particular range, and you don't have a way to calculate G7 based trajectories, I can certainly understand why you might do it this way. However, I cannot butcher the ballistics in that way for several reasons.
One; I simply know it's not the right way (after all, I paid a lot for my college education! Why would I blatantly do something I know is wrong?)
Two; my results have to be able to stand up to scientific scrutiny by my peers. If I took the 'two wrongs make a right' path, my methods would be identified as such and my peers would be the ones explaining to me on some internet forum what's wrong with my methods.
There are other details which are variations on the above two major reasons why I cannot reduce the data as you're suggesting. Again, if you choose to do so, feel free. I understand your reasons and in certain circumstances the error would be negligible enough that it would hurt you too much.
-Bryan
