Hi Ken
When studying the counter weighting applied to Craig's crankshaft it is apparent that the engine was considered as four 90 degree V-twins that happen to have their crankshafts connected. The advantage of doing it this way is that the loads trying to bend the crankshaft across the neighboring throws are greatly reduced. This advantage is probably more useful when considering the joints between the segments of a composite crank than it would be for one cut from a single piece of steel.
Another aspect that is evident is that the counterweights retain all of the original disc material opposite the crankpin from the line through the center of rotation. While material that close to the axis of rotation does not add much counter-balancing weight (hence the tungsten weights), this extra "flange" material increases the effective "flywheel" mass at each crankpin throw. In full size racing engines this extra flywheel mass would be cut away to improve throttle response, but in the case of a model engine with a composite crank, the extra flywheel mass at each crankpin throw usefully damps out torsional vibration that would tend to stress the joints between crankthrow components.
As for the total rotational inertia of the crankshaft, the additional inertia that results from the extra material at each crankpin throw flange can be subtracted from the flywheel proper, if desired. As Craig's engine will power a boat, building revs instantly would not seem to be a top priority. A slightly "heavy" flywheel will allow the engine to run smoothly at a lower speed than if the total combined crankshaft and flywheel inertia were to be reduced.
Consider that Steve Huck's original engine design retains the full crank flange discs and therefore the total effective flywheel inertia would be higher than for Craig's design interpretation (taking into account the 1.5:1 scale). Also, if Steve's engine runs smoothly enough without special balancing, then Craig's design can only make it smoother; especially at higher revs. Then again, Craig’s engine is larger and so are the loads that would benefit from balancing at any given rotational speed.
As far as the method for calculating counter balancing weights required is concerned, the link provided by Bluejets explains it rather well: http://modelenginenews.org/etw/etw_bal/p2.html