According to this link the beads move in the direction of motion which makes sense, however I don't see how they can move fast enough to counteract the vibration in real time. I'll stick with cheap and effective balance weights.
Let me first off say that I have not yet used their or other beads yet
(although I intend to try them at some point)
Their description could be clearer ...
Let me try to describe it better.
First of all, the best way to think of the beads is that they are a bunch of little balance weights that AUTOMAGICALLY move to the correct spot to balance the tire each time it starts to spin. The force that makes the tiny weights move to the right spot is the vibration of the imbalance itself. Their drawings try to describe the way the process works.
But, before we begin, a couple of assumptions that need to be spelled out
1) the inner surface of the tire is perfectly concentric to the axle
2) the inner surface of the tire is smooth enough that the beads can move relatively freely
3) that there is "more than enough" weight of beads installed
For the first drawing, we look at the tire sitting there not spinning.
All the beads are on the bottom.
For the second drawing, assume a perfectly balanced tire begining to spin.
Basically the beads spread around all around the tire and are evenly distributed. Once they are distributed, each bead basically stays in one spot on the tire as the tire goes around. (i.e. the beads are going around with the tire) ... It's like a tire with a whole bunch of tiny balance weights evenly spaced all around it.
For the third drawing, assume the same situation as the second, but add a heavy spot in the tire. As the heavy spot goes over the top, the tire "HOP"s up just a little bit (say 1/16 of an inch). The beads were originally evenly distributed around the tire (and spinning with it, not moving relative to the tire). As the tire hops up, inertia of the beads resists the hop, so the beads move 1/16 of an inch in the direction opposite the heavy spot and then continue to stay in the same (new) spot on the tire as it goes round. After a few hops, the distribution has changed to look like figure 3
and the tire is less out of balance ... The whole bunch of tiny balance weights are no longer evenly spaced on the tire, but are crawling around inside the tire to fix the balance.
For the fourth drawing, they are trying to show that as long as the tire has a little bit of remaining unbalance, the beads get a force that moves them to fix the unbalance ... the tiny weights keep crawling as long as there is imbalance
until in the fifth drawing the tire is balanced and all the beads stay in place on the tire as it rotates around ... all the tiny weights have reached their final position.
----
So thats how it works in theory anyways.
Does it work in real life too ?
It should; but, well remember those assumptions above ... If any of them are violated, then it won't work right.
If you don't have enough balls, they will all move to the opposite side, but still won't be heavy enough to offset the imbalance ... and I suspect that you need some excess just to have some distributed around the tire.
If the balls can't move around easily enough then they may never reach the spots that they need to. This point makes me wonder about the tire balance machine and if the balls can properly balance a tire in it. The machine may be too smooth, and some vibration probably helps the balls move around in the tire. Also, the suspension on a motorcycle lets the tire hop a little, the axle on the balance machine may hold the tire to rigidly for the tire motion to move the balls where they need to go. And finally, a tire on the road has the contact patch, where the tire flexes inward relative to its "neutral" position, which I'm sure helps the balls move around.
But the big thing that may cause problems is if the inside of the tire is not concentric enough to the axle. A thin spot could collect all the balls, even though it may not be opposite the heavy spot of the tire (that said, a thin spot IS pretty likely to be the light area on the tire and where the balls SHOULD collect anyway ... but its not
guaranteed to be the right spot)
I also found the MCN article a little suspect based on this sentence
Also, the weight of the beads added so close to the tire tread gave a noticeable increase in gyro stability, making the steering heavier
NUH UH, I don't think so ... 2 ounces of balls makes the steering heavier, but 1.6 ounces of lead an inch or two further in doesn't? (To say nothing of the several pounds of tire)
The two biggest selling points for the tire balls to me are:
1) Less work balancing (and you can't get it wrong by accident)
2) As the tire wears, the balance will change, and should be rebalanced every once in a while, with the balls you don't have to.
Well, that was way more than I expected to write at first.