This is what I believe at the moment :
MV is a measure of how consistent your air delivery system is, so shot to shot consistency and fps spread are quite important for this measurement. The actual number isn't as important.. so whether you get 770 or 790 doesn't tell you much (unless you're checking how close you are to the legal limit).
What shooters commonly do is measure MV with a combro for example, then plug that figure into chairgun using whatever chairgun's default value for BC is and use that trajectory to predict what's going to happen downrange. From dia6olo's test results, that's clearly going to be a recipe for lots of errors.
Downrange velocity is a measure of how efficient the pellet is in flight, the difference between MV and downrange velocity is the important measurement here - I'm not so sure how much relevance shot to shot consistency is. I would guess that if fps spread downrange is about the same as it is at the muzzle, then you can't hope for better than that really. The important figure though is the difference in velocity.
Downrange group sizes
This is an important test and for many is the only test they carry out. There are lots of factors other than pellet efficiency that can affect the group size, not least of which is the person pulling the trigger. I don't know about anyone else but for me, it's quite possible for me to shoot a really good group and a not so good group with the same gun/pellet combination. The difference is me!, so if I select pellets and rely on group sizes to make my selection, how do I know that the good groups that made me choose pellet A over pellet B weren't just down to me and not the pellet?
I think a better grouping test is to come up with what you consider to be an acceptable standard - say 25mm at 50 yards - and don't discount 'flyers' from the group size.
If you can consistently group pellet A and pellet B inside your acceptable group size, then don't pay too much attention to which one produced the tightest groups.
Why are there such large differences between different pellets?
I'm not entirely sure, but what is certain is that if 2 pellets leave the barrel at the same velocity but one is faster downrange than the other, the slower one is less efficient (and more prone to wind drift, will have a more pronounced trajectory and might possibly be less accurate). That might be because of reduced stability in flight or it might be caused by a less efficient pellet shape, or it could be some other reason?
Don't forget that a pellet changes shape when it leaves the barrel. There are rifling marks on pellets when they leave the barrel and pellets can also become squashed and sometimes elongated by the air delivery system, and that may well make a difference to the downrange velocity.
Is this all a bit anal and nerdy? wouldn't you be better served just putting in more practice?
Quite possibly. But... since I put that thread up some time ago about pellet selection with downrange chrono testing, I've had quite a lot of shooters approach me to share their own findings and what every single one of them has found is about the same as dia6olo's results - some batches of pellets seem to be significantly more efficient than other batches, and it doesn't seem to be at all related to the head size on the tin, in fact there doesn't seem to be a way other than testing the pellets to predict if it's going to be a good pellet or a bad pellet.
You can't tell by looking at the pellets or measuring them. (some dies produce longer pellets, and there are slight differences in pellet shape from one die to the next - shallow skirt / deep skirt for example).
So if you believe that (and the test results are pretty convincing!) then it makes sense to find a pellet through whatever means you prefer and buy as many of that die/batch of pellet as you can afford. (That's pretty common advice from most of the top shots in both FT and HFT).
If you're going to buy a lot of pellets, then it makes sense to do a bit of testing (however anal and nerdy that might be) before parting with a significant wedge of wonga.
Can you turn a bad batch of pellets into a good batch, by washing/weighing/lubing/sorting pellets?
I don't know.. my gut feeling is that some of those things might be better than others in some barrels.
The way to find out for sure though is by trying those things and doing some downrange chrono testing. I personally haven't seen or heard of any convincing test results, but the more people that try it the more data we'll have to help make a more informed decision.