The way that I figure it, is that the source of the vibration is the stylus arm going back and forth - so you want to stop that vibration as soon as possible, i.e. as close to the stylus as possible, before it starts the rest of the arm into motion.
If I understand your explanation, that's not what it's there for. The purpose is to increase the moment of inertia of the assembly of arm/cartridge so it has less propensity to rotate about its long axis. I would have to do some sketches but you might be right that lower it better. My gut feeling was that you want it right on the "flex axis" of the tonearm, which would be quite high, but that might not be the case (due to the parallel axis theorem).
Of course, making it longer helps a lot more (moment of inertia goes up with the square of the length, more or less). There's a point of diminishing returns, of course, and at some point it becomes too heavy. At least that's what my buddies and I found back when the Frank first described the longhorn modification. We made them of varying lengths and tack-glued them on (rigidly, but with 5-minute epoxy so we could get them off with a little heat). The overall performance was much better with the original short version, noticeably better still with it 50% longer, a little bit better with it twice as long, and then after that it didn't improve to the point we could notice. There were practical issues with it that long, too - the record wouldn't clear it when the tonearm was on the rest. It would just miss the spindle, but that wasn't much of a problem because you could bend it to provide clearance. As usual, we concluded that for a completely general application with no knowledge of the turntable, etc, that Frank's size was probably the best compromise between performance improvement and practicality. But for specific situations, you could get a bit more improvement if you made it as long as you could without running into something.
This was all using the original K&S 1/8" U-channel as the arm. It's clear to me that you could to a lot better as far as ratio of moment of inertia VS weight goes using something like carbon fiber thin-wall tube with something very dense like tantalum at the ends (and packed with clay for damping). Of course, you have to have both the tubes and the tantalum, which makes it, once again, impractical. I might dig out all that stuff and try it again, since it's now a topic of interest.
The *other* mods are at least as important, by the way.
Brett
