I don't think you'd see repeatable large sample data to show that weight sorting will make a difference.
For the 6.5 I'd buy better brass. ADG, Alpha or Peterson.
Then I'd shoot Sierra bullets. Sierra just seems to shoot with little fuss. Unless the gun doesn't like the powder.
My load dev is pretty simple.
Once brass is prepped with a bushing die.
Pick a bullet
Pick a powder that'll give max velocity and should shoot good.
Load 10 at book max with that powder.
Then I'll pick a second and third choice of powder and load 10.
Seat depth is a min of 60 thou. Sometimes more. Depending on magazine size. Form told us about going longer and the issues it can cause in the field. Once I set my seating did I forget it's adjustable.
Shoot single 10 shot groups with these testers.
Goal is 1 moa for 10 or a tad over.
If nothing works I may try a different powder or bullet.
I found that anything besides this just leads to a lot of tail chasing. It's 20-30 min for me to run out and test loads. Plus time to set everything up to shoot.
This is how I got this load for my 223 ai. A little different because I didn't know the max of powder. Tried a half grain under and it was good but slower. Added a half grain and no pressure issues and gave me the 2700+ I wanted.
I'll watch guys on YouTube chasing their tails on load dev. Trying powder charges by the half grain then seat depth etc. A half grain change won't have much effect on dispersion. So with three or five shots you're basically not seeing anything meaningful. You don't have enough sample size to see any changes you may have tried.
Not saying this is the right or only way. But for a blue collar dad with a bunch of kids it's the simplest way I've found.