This is true as far as it goes.
If you compare the predictive ability of Newtonian mechanics for the behaviour of bodies in motion, i.e. projectile flight, to the ability of an evolutionary biologist to predict any particular mutation or the evolution of any particular trait then the predictive ability of evolution, in terms of predicting the future course of evolution, is virtually non-existent.
I disagree. As I understand it Newtonian mechanics can't handle more than 2 bodies and be perfectly predictive. Evolution may be considered to be just as predictive but we have no "2 body" cases to deal with. They are all too complex to allow for total prediction.
It isn't a fault of the theory it is a challenge of the applications of it.
In fact, I may be wrong about the above. We have cases that are simple enough for pretty precise predictions to be possible.