Couple of things to try.
The most effective way is to increase substeps. If you're solving on 1 substep, it will have the effect of damping out the high frequency motion. Try a value of 4. It will be slower, but should give you better dynamics. Also be sure your tet resolution is high enough to support complex secondary motion.
There's also the stiffness damping attribute on the zSolver. This defaults to 0.01. You could try 0.001 or 0.0001. Something like that.
If you're trying to force more ballistic motion in general, you can try increasing the massDensity of the tissues (on the zMaterial node) Alternatively, if you wanted to see this effect globally, you could scale down the zSolver, which effectively makes the muscles larger in "solver space". Note that assuming your model is at the correct scale as it is, then you'd be moving away from 'real world' values if you adjust the solver scale, or mass density.