In my opinion if you are breeding for egg production then you need to track the average if a fair sized flock of the same age birds over a significant period of time say the first full year of laying.
Tracking by total and average by week and month.
Once you have these numbers then you can break each group of birds down into smaller breeding groups and determine how the group does compared to the wholes average. after 2 months of laying you should have a good idea if you have lagging birds you can then try trap nests or other methods to further break down a breeding group to remove the lagging birds.
You can use shorter time frames if you wish but giving the birds time to fully mature in their laying cycle before hatching would probably be good and should ensure larger stronger chicks.
There are so many things that can effect laying though including flock politics that moving groups of birds frequently can effect the numbers.
For example If I had a flock of 20 layers after determining my average. I would then sort breeding groups by pelvic widths and other possible SOP related characteristics into 3 groups of 5 birds ranking the groups. Then I would compare the 4 groups(including the left overs) to my previous averages for a couple of months. This should give you a good indication of how much uniformity there is in the flock as well as give you something some idea of what to breed forward from. By grouping around some uniformity standard this will insure that what goes forward should replicate. By using a largish breeding group (5 or so) you still maintain some genetic diversity
As for weighing chicks Id weigh weekly till culling or through 24 weeks for keeper birds as this would give me the best data about growth rates
Personally Id grow out a group of birds to 24 weeks so I could get as much data about how they finish as possible. This will also help select your forward breeders.
Feeding practices and a variety of other things can effect over all gain so I would grow them out in expected production conditions as best as possible. Feed input is an important factor to monitor. I would not change too many things at once that includes breeding select on. If you want to compare raising methods run multiples off the same breeding group. There are lots of factors that go into production qualities of the bird that have nothing to do with breeding. In order to ensure your breeding program achieves its best results then you need to ensure that you have a consistent method of raising the birds and maintaining their health. If you do that then you will get the best birds you can from the environment you provide. Once you have established a measure of uniformity from within that environment then the subtle changes to that environment can be easier to track. You have to determine how you plan to raise the birds. Commercial birds have been breed to be raised in an efficient environment for a short period of time. Your breeding program will produce the best possible from the environment and raising program you choose.