Your first sentence is a misconception. Yes, I've been testing things is a structured way for a long time, but I'm only providing numbers from Litz's "Modern Advancements In Long Range Shooting, Volume 3". If you only read one of them, read that one. The whole series is provided at their classes and you can often find sets for sale. That book also has stats for typical Velocity and BC spreads at the ELR match you're headed to. They're not nearly as rigorous as I'd expected. Knowing when you can stop running from the Bear and move on to the next thing is extremely useful.
According to that book, there is zero correlation between BC or Bullet speed and bullet weight. It's not a surprising result. The Velocity should go down, but not by as much as a quick kinetic energy calc would suggest because peak pressure will go up with a heavier bullet, then BC should go up with weight. They cancel enough to at least drop into the noise. One of the knocks on the ATips is their weight distribution is generally larger than the Hybrids. It's to be expected, the ATips are circumcised to add the tips and all those length differences seen in the Hybrids are snipped off.
That being said, I think there can be value in pushing the truly weird to the end of the box. This is a relatively common practice. If you have to adjust the shoulder bump for that batch of cases, the cases with multiple bumps are marked and go to the front of the box as foulers. Same with seating depth. Use the already marked cases to check and set the seating depth. Drop a case and ding the neck, hit it with the mandrel again and put it at the front of the box. Anything feels different on seating, front of the box. Spill powder dropping it in the case and have to recharge it, assume you didn't get all of the first charge out and put it at the front of the box. Primer seating feels off, front of the box.
On weight sorting in general, a tip to speed up the sorting and reduce errors is to work with just differences rather than 5 digits every time. Select one randomly and zero the scale with it. Set it aside. Weigh a few but don't start sorting yet. After 5-10, if there are about as many heavier ones as lighter go with that. Save the zero bullet off to the side to check scale drift. Now start weighing bullets and putting them in bins by 0.1 grain increments. I can do 5-10 /minute.
I fetishize case weight more than bullet weight. Case weight does not indicate case volume very well, but it doesn't take long and if you find the 1 truly off one, it'll save some mystery in your results. Checking case volume in new cases is pretty useless and very time consuming. Once fired is better, but only if all the cases have the same pressure history. Cases fired in new barrels or used for pressure ladders won't work. Measuring individual case volumes is extremely time consuming and prone to all kinds of errors, so buy a little extra good brass and use the ends of the weight sort for pressure ladders is my preferred approach.
So heavy to light or light to heavy? I usually load left side with the foulers and then light to heavy of whatever. What I try to do is use the ratty end of the box when it doesn't matter and the middle when it does.