[MikeK]

I guess (and I do mean guess!) it depends on where you want the crank to break. The traditional place is for the rear to snap off. Any piece of metal can only take X number of stresses before fracture. That number is a function of the degree of flex and the number of cycles. An 80 year old crank is already way towards that point. Adding throw counters and reducing the flywheel weight changes the dynamics of the shock waves, or flexes. You now have less at the rear flange and higher stress in other places. (Here a front dynamic element would be nice.)

As a result, you will get more remaining run time on an already 80 year old crank than would remain with the heavier wheel in place. Simply, you are gambling that your chances of crank failure will be reduced. Now, if you can find a NOS crank (keep dreaming) with zero flexes, you'll gain almost no confidence against failure unless you live another 80 years. It would be someone else's problem.

OK, now let's look at a NEW crank. Not NOS without counterweights, a new Scat, Crane, Burlington, etc. with counterweights. None (Well, maybe a $3K custom) is 100% countered for each throw. That crank will STILL always be doing a harmonics dance back and forth along it's length while running. ANY inertial dynamic introduced at either end will make the stresses unequal along the length. Now remember X number of stresses before fracture, that number a function of the degree of flex and the number of cycles. For maximum life of an I-4 flat crank, on each external end of the crank you need a dynamic(inertial) mass equivalent of sq.root of 2 (1.414) times the inertia of the adjacent crank element (throw). Nobody runs a front flywheel (solid damper) that big, and nobody has a rear flywheel that small. The best compromise is a rear flywheel as light as you can get. Of course, even with a heavy stock fly-anchor at the back, the average street A will never approach that cumulative X stress point within that owner's use time if you have a NEW crank to start. The only thing that owner will get is a marginally quicker acceleration.

Now, another consideration- If you reduce the inertial damping (actually, you are reducing the energy conservation of the shock wave) by lightening the flywheel, that shock wave continues down the drivetrain! No flywheel on the back would be like taking a #3 pneumatic rivet gun to the tranny teeth!

Everything is a compromise. Henry made his, based on NOS parts and then- conditions. There probably is no definitive right or wrong, it is only what you want to happen, or how you wish to prioritize and distribute probabilities of failure at various points in the entire system.