I appreciate your effort to create the perfect cast bullet wheelgun !!! There's never enough time or money, is there ?
Bearing length ? I don't have a formula for how long a bullet should be, but I can tell you that "decent" designs will have no more than 1 caliber of the length unsupported -- and by "unsupported," I mean the unsupported nose.
An example of a "decent" design would be Elmer's 44 bullet. I don't have one in front of me to measure, but I'm thinking the ogive was about 0.320" long, maybe a little longer, making the ogive length about 3/4 of the diameter. I personally think Elmer's ogive is longer than ideal. He may have done that on purpose because he was keen on long distance stability and because he needed more powder space with the 44 special case and the powders of that era.
Another example of a "decent" design is Veral's LFN. Its unsupported nose is what -- about 0.300" long ??? Like Elmer, Veral advocates using a long nose so you can stuff more powder in the case.
At the time Veral developed his designs, pressure testing equipment was not available to the common man. We now know that, when loaded to equal pressures and shot out of a "real" handgun that you can carry on your hip all day long, deep seated short nose bullets produce almost the same velocity as shallow seated long nose bullets (except in the case of very heavy bullets that cramp the powder space). So with a few exceptions, I don't buy the "powder space" excuse for long unsupported noses.
On "better" designs, the unsupported section will be much less than 1 caliber long, more like 1/2 - 2/3 of a caliber long. This is true even for cast spitzers that are used in 200 yard benchrest competition. You never, ever, see long unsupported noses winning cast benchrest matches. Never.
An example of a "better" design is the Eagan target bullet. Again, I don't have one in front of me to measure, but from memory, the unsupported ogive is only about 2/3 caliber in length. That is on a bullet that is over an inch long. Another way to look at it is that about 85% of the bullet is supported.
I don't worry about lube grooves when calculating bearing length. Lube grooves don't seem to create any support problems.
So to sum things up, I like to see 75% or more of the bullet supported, and the unsupported nose should be no more than 2/3 of a caliber long.
Ross Seyfreid once said something to the effect that "the ballistic coefficient of a cast bullet is zero." In other words, don't worry about aerodynamics. A hunting bullet has to have a minimum of aerodynamics or else it won't stabilize, but otherwise, don't worry about it. It's certainly desirable for a 200 yard benchrest bullet to have good aerodynamics, but even so, you don't see VLD designs winning CBA matches. They use spitzers, yes, but abbreviated spitzers, with a short ogive. Support is more important than aerodynamics.
Getting back to your revolver...... you seem to have given this a lot of thought and there is not much I can tell you that you don't already know. But I do challenge your statement that a short nose has to "jump" further. Not necessarily true. The thing that has to "jump" is the front band. The unsupported ogive doesn't "jump" because it never touches the rifling. It is the length of the front band that determines how far the bullet has to jump, not the length of the nose.
You can get around this by using a long undersize front band, or else a long bore riding section. Say you make an extra long front band or bore rider 0.448" - 0.450" diameter. That willl drop into the chamber easily, yet it is still fat enough to touch the rifling. I can buy into that as long as you are realistic about tolerances. I don't think 0.0005" clearance is realistic. More like 0.002". As you have noted, if clearances are too tight, you can run into intermittent chambering problems due to residual lube or grains of unburned powder or just normal bullet tolerances.
I'm totally with you on the weight thing. I tried to get 240-250 grainers to shoot in my 44 from hell, but was forced to go to a 280 to get decent accuracy. I tried to get a 180 to shoot in my Marlin 357, but was forced to go to a 190 to get decent accuracy. I tried to find a good 165 grain load in my 30-06, but found that 180's were much more accurate. And so forth. Longer bearing length and additional lube grooves seem to fix a lot of problems.
The only other thing I can suggest is to find and read Ross's article on the paradox revolver, or any of his other Paradox articles. I think that was in Guns and Ammo, many, many years ago. Ross had a revolver made to shoot both shotshells and cast bullets. He had the rifling cut out for all but the last inch of the barrel. I think it was a 7 1/2" barrel with rifling only for the last inch, or something like that.
You may be wondering, "what does that have to do with me? I don't give a rat's ass about shooting shotshells." Ahh, but all revolver bullets have to "jump" before they touch the rifling. The Taylor throat has to jump a little farther. What Ross did was make the world's longest Taylor throat. It made the "jump" problem much, much worse, and he had one heck of a time getting it to shoot cast, as you can imagine.
But Ross found a solution to his "jump" problem. Do you think his solution might apply to our "jump" problem, as well ? Maybe.
Ross's Paradox article never got much attention, because only two or three people in the world care about Paradox revolvers. But If there were a Nobel prize for cast bullet research, Ross should get it just for that one article.