Hmm. I've never experienced any accuracy degradation from a gas checked bullet extending below the neck and am inclined to say that's a myth unless there is another factor at play of which I'm unknowingly avoiding somehow through my loading techniques.
I also tend to think that a long neck alone isn't going to do much to support the bullet after the primer goes off unless the neck is thick, hard, and has a fire-formed ring at the base and enough flare left on the mouth to scuff the chamber neck...and also the launch pressure peak is less than that required to blow the neck out to the confines of the chamber neck walls. A thin .30-30 or .39-40 neck? Doubtful it supports the bullet much in the middle of the empty space around itself in the chamber.
I'll also offer for consideration the loading technique which John Ardito used to such good effect in competition: Throating the rifle to take almost all of the bullet and leave only the gas check in the case neck for handling; a technique which is about as close to breech seating as possible with fixed ammunition. If using the brass to support the bullet in-line with the bore was of any great accuracy benefit, wouldn't he would have used that method instead to win matches?
My personal theory on the long neck belief is that most of the "long neck" cartridges are from antiquity and are from low-pressure cartridges which require very little chamber neck clearance for safety in releasing the bullet. In modern, high-pressure bottleneck cartridges there is often .008-.010" or more total loaded neck clearance with a groove-diameter bullet. Compare that to a .30-30 and SAAMI drawings and then add a couple thousandths to the loaded neck diameter when using cast bullets and the tolerances decrease substantially....to the benefit of accuracy.
Consider the 222 Remington, topic of this thread. SAAMI indicates a maximum loaded cartridge neck diameter of .253" with a minus tolerance of .008", and a chamber neck of .254 X .255" with a plus tolerance of .002". A max cartridge and minimum neck would actually have .001" interference at the case mouth.
The .30-30 Winchester has a maximum loaded cartridge neck diameter of .3301 X .3331" minus .008" and a minimum chamber neck of .3307 X .3337" plus .002". Again, there could possibly be interference with minimum chamber maximum cartridge, and if dead-on spec for both is achieved, a half-thousandth TOTAL neck clearance will exist. Most .30-30 brass is about .0115" to 0120" thick giving up to a .334" loaded neck diameter with a .310" bullet. Easily an interference fit with even a maximum chamber....and most people load .311" bullets in their .30-30s. Short version....IT'S A VERY SNUG FIT with the case neck and chamber neck taking up most of the space where the back half of the bullet could possibly get crooked at launch. Launching bullets straight into the bore is one of the essential keys of cast bullet accuracy.
Now, compare those tolerances to the .308 Winchester cartridge. The SAAMI .308 Win cartridge neck maximum is .3435" (no taper) minus .008" and chamber of .3442 X .3462" plus .002". Still possible for a .0003" interference, but in reality necks are no thicker than .014" for a loaded diameter of .336" with a .308" bullet and .338" with a .310" cast bullet, leaving .006-8" total clearance with a .310" cast bullet. It would take a .315" bullet to get close to the same fit with a .308 Win as with a .310" bullet in a .30-30, in other words A VERY SLOPPY FIT with typical components and chambers.
Again, I believe it's this tolerance reduction which makes some of the legendary cast bullet calibers work rather than their typical long necks.