I think the "longer bullets are inherently more accurate" stuff comes from the basic idea that longer bullets have more surface contact with the rifling in the barrel, or have three driving bands instead of two, thereby imparting a bit more stability as the slug travels down the barrel. Another idea stems from the connection between bullet weight and the rifling rate for a given caliber. This latter is the heart of benchrest rifle accuracy, but it applies to pistols, too. There certainly are more than enough theories about revolver accuracy and certain revolver/bullet/powder combinations to keep hobbyists busy for years. At this point in my shooting, if I can keep 90% of my .44 magnum rounds on a 8 inch target center offhand at 25 yards, I'm happy.
Thanks for the reply. I was aware of the relationship of twist rate and bullet weight (a non stabilized bullet is an ugly thing). All other things being equal, (cardtridge, bullet, COL, primer, velocity) powder type is crucial. Nothing I've used beats IMR, and when I have called Remington, Spear, Sierra, Nosler, etc to pick their brains for an accuracy load, they have never recommended anything else.
The Bench Rest folks are like racers. They leave no stone unturned in the quest for better performance. Many conversations have turned to how many fairies can dance on a pin head I'm sure.