His reference to the shrimp's disability is mindful of the theory that bats evolved from rats via a gradual elongation of the medial digit of the forelegs, over zillions of generations, until they finally supported flaps of tissue as the leading edges of two useful wings. The problem with this theory is that the intermediate stucture which would characterize half a zillion of these generations would be neither a useable wing nor a useable foot.
And if natural selection is as powerful an engine as the evolutionists say it is, the half-zillion generations would provide more than ample time for natural selection to kill the crippled animal.
In other words, natural selection tends to prevent bridging from one (existing) species to another (prospective) species via mutation. Vaguely invoking plenty of time and astronomical numbers of one sort or another won't really help. The universe isn't old enough for all of this stuff to have happened. The bat-to-rat difficulty is only one such problem. There are billions more like it.
And the monarch butterfly really is completely inexplicable by standard Darwinian theory. It always will be.
(These problems constitute the reason why Stephen Jay Gould has abandoned standard Darwinian theory in favor of his "hopeful monster" theory. He finally noticed that the creationists' complaints concerning the standard theories of mutation and natural select are crushingly serious.)
Besides, the insurmountable problem concerning the mutation of rat toes to form bat wings also fits the more general complaint by creationists to the effect that mutations are virtually never benefic. They do disable.
I think an engineer would not say concerning the rat-to-bat theory "Well, that's what must have happened." Rather, he would model the scenario of interest using some best-case probabilities and say, "We cannot even say it might have happened. What we need to say is that it definitely didn't happen!"
You are making WAY too much sense to be included in OUR C vs E 'debates'.