Accuracy in a handgun is subjective to your proficiency, how you use it, how many rounds you use and if (or what) its being measured against.
Different disciplines have different benchmarks.
I primarily do action pistol shooting (IPSC) so when I develop a load for my pistols my accuracy goal is to reliably hit a 6” plate at 20 yards (slow firing) and to maintain 90% A hits at 10 yards fast firing. It must make power factors, it must be reliable and I have to be able to mass manufacture it.
To attain my goal I accept that half or more of the results on paper are my errors, not the loads.
So to start I do the research to see what everyone else is using and what specifically works for them in my gun (OAL can vary along with other tweaks). The wheel doesn’t have to be reinvented but don’t blindly trust what you read on the net although patterns do emerge.
This info usually isn’t a secret and eliminates lots of wasted time and money.
Once the components are secured I make small batches to chrono (must make power factor) all the while keeping a close eye on function. The final load must be 100% reliable.
Once reliability and velocity are down I shoot for groups as I would normally use the pistol.
If the groups don’t provide what I expect on target I will adjust the powder and/or OAL, in some extremes a different powder or bullet weight/style.
I also keep an eye on how dirty the load is (action and barrel), some combos are very dirty, it’s no fun to gum up your pistol halfway through a match so why not develop a cleaner loading.
All that said, I find load development on a rifle is more scientific/OCD and on a pistol more of an art or by feeling (Felt recoil is real).
A rifle I can spend hours with few rounds completed (looking for a tack driver), a pistol load must be able to be mass produced to be successful (how I use it anyway) so some long processes to try and make a tack driver just don’t cut it.
I find it takes some time to find a load for a pistol and that final testing must be done under the same conditions that you intend to use it in; if you dynamically shoot then do final testing dynamically, if static, then do the same.
