BHuij
So this mostly matches what I understood about twist. Namely that an uneven, rough, wrinkly, or otherwise defective boolit will destabilize at a much lower RPM than a smooth and even one. However, I also understand that boolit length/weight, diameter, and alloy (or more directly, hardness) has a lot to do with the RPM at which the projectile is most stable. For example, a 77 grain projectile (which is inherently longer than a 55 grain projectile) will need a faster spin to enter the "stable" zone, as well as a faster spin to exceed the "stable" zone and lose its accuracy.
Your understanding about twist and RPM is incorrect. An "uneven, rough, wrinkly, or otherwise defective boolit" will not destabilize, "at a much lower RPM than a smooth and even one", any bullet not even a cast bullet. If the bullet has sufficient RPM for stabilization it will be and will remain stable regardless of the bore condition. Also, there is no "stable zone" once the bullet is stable it remains stable. Over spinning a bullet does not cause it to lose stability. Over spinning a bullet with any imbalance can and does cause it to lose accuracy by causing the bullet to delineate from its intended flight path. The farther the range the greater that delineation will be.
The RPM required for stabilization and the adverse affect the centrifugal force has on imbalances in the bullet caused by RPM are two completely different things and affect the bullet in two completely different ways. As mentioned, a perfect bullet coming out of the bore stabilized will shoot very accurately, basically into one hole shot after shot. The trick is casting and getting that bullet out of the bore perfectly balanced. That is what match shooters, particularly bench rest shooters including cast bullet bench rest shooters, strive fore.....with both cast and jacketed bullets......it is their holy grail. As I said, once the bullet is stable it will remain stable across most if not all of it's flight path.
That leaves the centrifugal force aspect of RPM that is causing the larger group sizes.
Now, that being said, there have been numerous threads regarding the adverse affect RPM can have on bullets, especially cast bullets, at certain levels of RPM. No need to rehash that here but it is suffice to say that you are actually seeing the affects RPM is having on your bullets, i.e. the large groups. If you really want to find a useable cast bullet load for your 9" twist AR that will consistently group into 2 to 4 moa with the 55 gr cast bullet you are using then I suggest you use an easily ignited powder such as H322, RL 7, 3031 or H4895 that, when used with a Dacron filler, will function the action reliably while maintaining a velocity in the 1850 to possibly 2250 fps range. You can choose to not believe what I just said but unless you can cast perfectly balanced bullets and then get them out the end of the ARs barrel perfectly balanced you will find it so........because you already have.
Let me add when I say "consistently group" I am not talking about a single 3 or 5 shot group at 50 or even 100 yards. I am talking about at least several magazines full over a days shooting, especially with an AR. Being able to shoot a 3 or 5 shot group with any bullet means nothing to me with an AR. The AR and its ammunition must prove reliable and consistently accurate for many rounds or it is useless, at least to me anyway. I am also talking about shooting to at least 200 yards if not 300 yards with a cast bullet "practice load". However, if you are interested in only 50 or 100 yard max range, as many are with such loads, then you can boost the velocity to 2400+ fps and maintain 3 - 5 moa accuracy over several magazines full of such a load. However, you must cast and load a more perfect bullet even then.