I'm curious how precisely close each injector squirts fuel in relation to the other injectors? I know there are certain tests you can do on injectors but I've still wondered do they all squirt exactly the same throughout the RPM range. Similarly is the volume of air that is injested exactly the same based on the shape of the air box, spark, etc. I do understand what rivers is getting at though. Why not be as precise as you can with the tools you have on the things you can control? I wish I had a set of feeler gauges with more granularity but I haven't been able to come up with a set so I can only get "close" with the set I have, and that's only if I am using them correctly. Measuring tolerances definitely takes some accurate measuring tools, knowledge, experience, practice, and repeatability. A buddy of mine was talking about how you can change valve timing based on the difference in how you adjust within the tolerances between the intake and the exhaust valves. Setting the intake valves to the tight side of center and the exhaust valves to the loose side of center for example and how that can effect duration and performance and why.
It all comes down to how finely do you want to split the hair -- there is always more that can
be done; it comes down to how much return is there on the investment. I too, get what riverszzr is saying... the answer for me is that the manufacturer has a lot more time and effort sunk into deciding that than I ever will.
Injector timing? Does the manufacturer account for the length of the wires being different between injectors from the ECU? The signals may arrive at the injector at different delays as a result -- this is something that is a concern in circuit board design.
Signals propagate through copper at about 1 picosecond (1x10^-12) an inch. That sort of accuracy probably isn't required, when the rate the injectors open and close may vary by a far greater amount. How exactly does that matter? Without a LOT of research and actual measurement, it's a guess.
Getting the ECU programmed more accurately than the mfg did it, is probably the fastest way to more powrt and better performance. As the end use can ignore all of the environmental BS imposed on them, the charges can be richer, timing more precise, etc. etc.
Indexing cams so they have the best possible overlap, duration, and timing are all desireable things -- I'd expect a few percent change in HP at the rear wheel are possible.
Air flow and resonances in the airbox, intake, exhaust.... the two ways to sort that out are directly related. Trial and error got us from side valve, to overhead valve, to overhead cam, to shim under bucket valve control. Somewhere in the middle of that, when the processing power got high enough, computational fluid dynamics has come into it; those computer models are by this point probably pretty good at estimating how much power a given configuration may produce.
The model will still not exactly match what happens on the dyno -- and that will still not match what happens in actual use. the differences are smaller than they have ever been. But they are still there.
There will always be room for improvement.....