I've been reloading for a long time, there are no short answers to your question. I could say something esoteric like "it takes what it takes" but all that would prove is I can pi$$ you off.
Here's my first tip: stay on the page of a good reloading manual. They typically give a range or max and some increment to deduct for a starting load. In general this range of weights is in a well behaved region of the powders capabilities and everything should be good (all should cycle the gun and deliver the nominal accuracy specs).
I bought a chronograph early in my reloading and use it extensively, but others swear buy visual inspection for pressure issues and actual groups on target. Here's where it gets 'dicey'. You must know what the gun is actually capable of shooting (or what claim of capability) from the maker. Most OTC stock rifles are 1 to 1/5 MOA, Most OTC stock defensive handguns are 3 inches at 25 yards, but there are variations. Key, of course, is that you can actually shoot to these levels, you have the proper support and a measured distance that is level.
I can and I have the necessary tools. As a preliminary 'cut' I take the top powder weight, subtract the lower powder weight and divide by 4; this gives me five weights to try. I load ten of each weight and head to the range. I chronograph five rounds and shoot five rounds for a group. I look for a small spread, a small standard deviation and a tight group; often I get two out of three, on occasion I've gotten two sets where all three parameters are good (good is relative--better than the other rounds at other weights is more correct).
I then load the best performers and check +-1/2 of whatever increment I got in the previous math--3 weights, fifteen rounds each. Three go over the chronograph again, 10 go into two groups of 5 at either 25 yards or 100 yards. Two are for alibi's, if all the holes are nice and round otherwise a stability check must be performed against a large poster board target at intermediate distances and beyond the stated standard distances.
Handgun ammo must function in the gun, if it's for SD, practice or games it usually must meet a power factor, it should shoot 4-10 rounds into 3 inches at 25 yards at will from a rest and it should be stable at 50 yards (doesn't tumble and show key holes). Rifle ammo must cycle reliable and easily for hunting, self defense and plinking, it must be predictably accurate at range and it too, must be stable beyond the maximum anticipated range. In some calibers there is a danger of premature bullet fragmentation-light construction high spin---you know as soon as you fire.
I buy at least 300 rounds of anything new, more typically I acquire 500-10000 rounds (all components) for established loads.
TO be all inclusive I see you are referred to both Dan Newberry's Optimal Charge Weight Method and the Audette's Ladder method. It takes a great deal of courage to put ones personal work up for public inspection and review and that is to be admired.
It's important that we see things for what they are-and that's a safety issue that must take precedent. Newberry's OCW is just a fancy way to 'hopefully' reduce two common data collections errors: systemic error-increased powder weight SHOULD increase velocity and operator error-one tester with one metric is insufficient for statistical completeness. And all the instructions, commentary and innuendo are smoke to cover the fact he is a cheap bastard and doesn't have the funds or the will to spend the funds to produce exceptional ammo and he likes 'dissing' those that do.
Audette's Ladder is specifically for well constructed and highly 'tuned' match rifles at 200+ yards. While it's grounded in sound science, it's pure hell on the 'pair' that undertakes conducting the test. Or you will spend a bunch of money getting the best optics you can find. Two key points: the gun must be capable of sub MOA grouping (think bullet diameter plus a few thousandths) and the shooter must be able to perform to the same level----that might be 3-4% of the shooting population.
