Balancing
It seems like nearly everyone has an opinion on engine balancing and the ideal balance factor. Search the internet forums for "engine balancing" and you'll soon be wading through thousands of posts. Unfortunately most of them will contain nothing but misinformed opinions, hearsay, old wives tales and plain old BS. Pretty typical misinformation levels for car forums come to think of it. For what they're worth here are my opinions...
Depending on the application, balancing can be either a non-issue or it could be of supreme importance. Mild street engines with a rev ceiling of not much more than say 5000 - 5500rpm will survive quite nicely without any particular attention, and with either a counterbalanced or non-counterbalanced crank. Providing the piston and rod weights are reasonably matched I wouldn't bother with balancing at all. If it makes you feel better though, go ahead and balance it.
As the rev range rises though, it becomes more and more important to both use a counterweighted crank and to have the entire rotating assembly balanced. Once you get into the speed ranges of 6500rpm and above, it becomes critical to use the lightest possible rods and pistons along with a fully counterbalanced crank in order to avoid breaking blocks.
Before we go any further we might just take a look at how the rotating and reciprocating masses act within a multi cylinder inline engine. As usual this'll be a grossly oversimplified explanation but hopefully it will help. Picture a typical inline four (it's simpler than a six but the same principles apply). It will have a single plane crank, where pistons 1 and 4 rise and fall together, as do 2 and 3 which are 180 degrees from the end cylinders. Ignoring the rotating masses for a minute imagine pistons 1 and 4 approaching TDC - by the way the crank on our engine has no counterweights whatsoever. As they are slowed down they pull up on the rods and this force is transmitted through the crank to the block. Now, if there were no counteracting force applied these two pistons would be jerking the engine up and down every time they pass through top or bottom dead centre. Riders of old Triumph and Norton twins will understand this intimately. The thing is though, as 1 and 4 approach TDC pistons 2 and 3 will be slowing as well as they approach the opposite dead centre, providing a countering downward force that exactly cancels out the force from the other two cylinders. Similarly, if we now look at the rotating masses (crankpins, rod big ends etc.) we see that they also balance each other out (being 180 deg apart) without needing any counterweights on the crank at all. The Holden six has a similar inherent balance, and so runs more or less free of vibration even without counterweighting - and even if it does have counterweights the vibration level won't change much regardless of the balance factor used. (In practice, the Holden six can buzz quite badly at high revs. But this is mainly due to torsional vibration of the crankshaft, a different kettle of fish altogether and mainly unrelated to balancing).
Okay then, if vibration isn't a problem why is it so important to run a counterbalanced crank on our sixes? Let's go back to our four cylinder engine for a minute. Imagine the crank is spinning at say 8000rpm, and pistons 1 and 4 are being slowed down as they approach the top of their stroke. Lets say each of these pistons is applying a force equivalent to about 2000kg, and likewise pistons 2 and 3 are each applying the same force downwards. Obviously the forces will cancel each other out and there will be no tendency to vibrate. But look at where the forces are applied. In effect we have a force equivalent to 2 tonnes applied upwards at each end of the block, while another 4 tonnes is applied downwards to the centre. In other words we are expecting the block to act as a beam as we apply pretty severe bending loads to it. Imagine supporting the block at each end and pushing down on the middle with a hydraulic press, then releasing the weight before rolling it over and repeating the procedure. Now imagine doing this 16000 times a minute. Looked at in this way it's pretty easy to see why the Holden blocks tend to disintegrate when subjected to high speeds with a non-counterweighted crank. Remember too that these huge loads make their way to the block through the crank and the bearings, so these too are stressed considerably.
It's pretty obvious that if we fit counterweights to each crank web we can provide a counteracting force for each cylinder that acts on the individual cylinders axis, and thereby eliminating the bending loads on the crank and block. If we use a balance factor of 100% - ie. a counterweight equivalent to all of the rotating mass as well as all of the reciprocating mass - then the loads on the main journals and block can be completely relieved at top and bottom centres. This introduces another problem though. At mid stroke the counterweights will be applying lateral forces at opposing locations to the block, so we still have bending forces at work. The solution is to compromise by using a balance factor of around 50% instead - ie. a counterweight equivalent to all of the rotating mass plus half of the reciprocating mass. This way we reduce the loads at top and bottom centres by about half, while introducing lateral loads at mid stroke that are of a similar magnitude. In other words we swap two big vertical forces for two small vertical forces plus two small lateral forces. This 50% balance factor will provide the smallest possible loading at any given point in the cycle, and should be used by default for most engines. It seems fashionable at the moment to "overbalance" engines for very high rpm work, using balance factors of 60 or more percent. Typically though, the people doing this provide no logical or convincing reasoning for this. I guess if you had reliable data showing that your block is stiffer laterally than vertically, or vice-versa, then you may be able to justify some amount of under or over balance. In the absence of this data though I'd stick with about 50%.
Some builders like to add a few grams for oil, though how the hell they know exactly where the oil will be clinging to the rotating and reciprocating bits is beyond me. Some also painstakingly balance everything to a jillionth of a gram but like the oil thing this is just wank. If the individual components share weights within a few grams, and the balance factor is somewhere near an appropriate figure then it's as good as it's going to get. Taking it to ridiculously fine tolerances will achieve no measurable results.
We briefly looked at knife-edged cranks in the crankshaft section, but it may pay to reiterate here: if engine durability is a factor do not knife-edge the counterweights. It will be impossible to balance one of these cranks to a 50% balance factor. Your engine builder may tell you that he can balance the modified crank perfectly (and he probably can) but it will end up being balanced to something like a 30% factor. For a high revving Holden six that's already somewhat fragile I feel the slight gains from reduced weight and windage are far outweighed by the significantly increased stresses introduced by the cut-down counterweights.
So what do we do if we want or need to rev our engine to the moon (and remain in one piece)? First of all we minimise the forces by using the lightest pistons and rods available/affordable. Then, by using a fully counterbalanced crank we can reduce the abuse the block has to absorb by about 50%. A heavy, rigid main girdle can also help by dampening vibrations and stiffening the block laterally. Builders of smaller, 3" stroke engines are disadvantaged in that counterbalanced cranks aren't available. In the Holden sixes heyday engine builders recognized the importance of counterbalancing and went to the trouble of welding weights to the original steel cranks, though this will obviously be time-consuming and/or expensive. A billet crank is another expensive alternative, otherwise you're pretty much stuck with using a late 202 crank.
Now the bad news: with conventional balancing equipment it simply isn't possible to balance a straight six crank to a specific balance factor. If the crank is balanced end for end the balancing machine will show no change in balance regardless of what bobweight you use. The only practical way to do it is to carefully measure the crank throws and calculate the factor theoretically. This isn't as difficult as it sounds; CAD software can do the sums for you. And if you plan to use a billet crank then the design software will give accurate counterweight dimensions for a given bobweight. In the good old days people would cut up an old crank and measure the balance of each throw individually though today that would be a last resort. The main thing is to be aware of the limitations of the commonly used balancing machines.