The percentages you keep seeing represent nothing more than a % value given to the amount of reciprocating weight of the piston+rod+rings+pins in the engine.
To put it simply, when they throw a V6 crank on the balancer, they have to simulate the weight of the piston, rods, and associated parts. Since some of that stuff is rotating, and some of that stuff is reciprocating, engineers have come up with this idea to assign the reciprocating components a % value that is less than actual.
Buick V6 engineers weighed the piston, pin, small end of the rod, rings, and even assumed a small amount of oil and then decided to multiply that by .366 which then gives you a value that is 36% of actual weight. The rotating half of the rod which includes the bearing is factored at 100%.
These two weights added together give you your "bobweight" which is just a deal that clamps on to the crank rod journal that the machinist adds or subtracts weight with small washers (from the videos I have seen online).
So the higher percentage of weight does not really have anything to do with how close the tolerance is. Small block chevys for example are typically balanced with a 50% balance factor for the reciprocating components.
What a lot of people don't realize is that this balancing business is really quite complicated. I don't think most people that do engine balancing have a grasp on what's going on. (I certainly don't either, I have more questions the more I learn about it) Consider that the force exerted on the piston changes its "weight" in a running engine, on the power stroke it's pushing on the crank, and on the intake its pulling, or that a crank counterweight is not moving in the same plane of rotation as the forces exerted at the rod journal. There are about a hundred other things at work.
The basic gist I've gotten from this all is that the % given to the recip. weight changes the plane on which the imbalanced forces act. From what people say, 36.6 moves these laterally which allows the engine mounts to soak up felt vibration more effectively in the buick V6, but that the imbalanced forces are greater than if at 50% which moves the forces vertically. It would seem to me there is a side benefit in that moving the plane of imbalance can maybe move the forces in a direction that is stronger in the block. Maybe these main web cracks people get is partially due to lateral forces from the crank.. who knows.
What I do know is that the information on this is scant. In my research I could not find any info that directly showed the imbalanced force direction, it's magnitude, or anything like that. Just second hand info.
For my own build I used a stock crank, balancer, flex plate, rods, and std bearings. I did however use TRW pistons and instead of using the heavy piston pin they come with, I put in stock pins. This made for a much lighter piston and pin combo than stock. I had these parts match balanced to each other and that was it. I did not have the crank balanced. The point being that my change in the static imbalance was nothing more than effectively increasing the bobweight% which I felt was desirable anyway. I think I worked out the math to show that it was equivalent to like a 42% on the bobweight. The dynamic imbalance is the same if not better than stock now since the rods/pistons are matched to each other to a tighter tolerance than stock.
The result of this insanity? My engine is just as smooth as it always has been, hasn't blown up, and just feels great.
Just something to consider. Read the wiki on balancing and the articles you can find on google. Lots of interesting stuff.