Does option "alpha" in nonlinearmodel.coefci still work?

Following this question, I'm interested in obtaining confidence intervals in the parameter my nonlinear fit of a sin curve. I see that these confidence intervals for parameter associated with the phase of the sin are not just extremely large (ranging from pi to 2pi), but they appear to not change with different choices of alpha.

coefCI(myFit_out, .99) and coefCI(myFit_out, .01) both return the same set of values. Is there some sort of error or issue with my utilization of this?