There has been published empirical research by Alan Carruth into break angle and string height.
His experiments find that break angle makes no measurable difference in volume or wave form within a broad range (from memory, 11 degrees seems to be the minimum required), though a higher break angle increases the tilting force on the saddle and thus increases the risk of breaking out the front of the bridge.
By contrast, changing the height of the saddle peak above the soundboard does change volume measurably.
The physics of all this is complicated and somewhat counter-intuitive - for example, the back and forward movement of the saddle doesn't contribute appreciably to the fundamental note but does modify the harmonics of the string, and the soundhole does not 'let the sound out'.
The standard dimensions and angles traditionally used for stringed instruments seem to fall pretty much at the optimum values. This shouldn't be surprising (though it is for those of us who are used to scientific advances) because traditional instrument design is the product of centuries of trial and error, and ought to have been expected to converge on the best solutions. This doesn't mean that new ideas aren't worth trying, but I suspect that changing just one aspect of design (eg a new kind of bridge) while keeping the rest of the instrument the same is likely to produce an instrument which works less well. Radical redesign of the whole (eg resonator instruments, electric guitars) is a different matter of course.