on shunt up to 500Vac. Do you have more preciese phase shift data, than that
shown in Figures 30 and 31 in Datasheet?
For your application "line voltage monitoring at 50/60Hz", I think the phase
shift can be ignored. From measurement, it looks like the input/output delay of
Figure 29 is several microseconds, so I think it is hard to observe for an
input signal at 50/60Hz.
Figures 30 and 31 are used to sketch the bode plot and demonstrate why we need
an integrator connected for the close loop test.