Well, I've done a bit(?) more Googlin' on the subject of balancing / over-balancing and even under-balancing. David Vizard seems to have gone from one recommendation to another lately, since his latest tech stuff mentions how Richard Maskin told him how they (Dart) found some quantifiable improvements from under-balancing some engines. Apparently DV's latest approach is to have his BB (doesn't specify how big) cranks underbalanced 1.5%.

So... short of a fully-funded research program using multiple engine configurations and a wide variety of design parameters, I'm back to the idea of going for a traditional 50% (and don't see a need to get all worked up over exactly how accurate that is). IMO, doing anything else falls into that broad category of "It depends..."