Although I'm not a math whiz, I can follow it well enough to see that the main signal the strings produce depends only on the string tension and the angle the string is displaced through before you release it during the pluck. Also, I've wasted 'way too much time measuring the forces a plucked string generates at the saddle top, and the data points to the same conclusion. If you push a .014" high E string down by, say, 1/8" at a point 5" away from the bridge, and let it go, the forces it produces at the bridge top will be the same no matter what the break angle is, or what the bridge is tied to, for the most part. The only exceptions I can think of involve problems, such as not having enough break angle, so that the string hops off the top of the bridge during part of the cycle. Since the string is seldom, if ever, displaced by more than five or six degrees at the bridge when it is plucked, it would seem that we don't need much break angle to prevent string hops.
Another problem is 'wolf' notes, where the bridge moves so much at particular pitches that it no longer presents a 'stop' for the string that is firm enough. In these cases the length of the string is not totally defined, so the pitch is not either, and the energy tends to be extracted from the string too fast to be of any musical use. One can see how a large downbearing force at the bridge would serve to 'nail' the top and prevent wolf notes, but it can also kill any chance of the top vibrating.
It's possible that there are certain geometries involving the top arch height, neck angle, bridge height, and so on, that serve to strike the 'right' balance between too much bridge motion that yields a wolf, and too little, that produces no sound at all (a la Les Paul). I've seen diagrams of violin setups that claim to show this (usually also involving the sound post, which we don't have), but I'm skeptical.
If this _does_ work, the only mechanism I can think of would involve a balance of forces. String tension puts a compressive force along the length of the top, which, by itself, would tend to make it belly upward. It's possible that having an equal countervailing down force provided by the break angle over the bridge would serve to free up the top to move. Since both forces would be proportional to string tension in the same way the 'best' break angle would be determined by the geometry. One outcome of this balance would be that the bridge would not be displaced either upward or downward when tension was put on the strings. Thus it would be fairly easy to test the idea out. Naturally, I haven't done it yet!
If I'm right about this, then there is a 'proper' break angle for any given guitar, which may well be pretty high in some cases. This does not, however, translate into 'more is always better'. That is the argument of the idiot with the pepper shaker, and I don't think any of us are idiots.
