And how should we tell whether Na and Cl are ionized in solid NaCl? Is this question any answerable?
Since every Na is surrounded by Cl atoms that "touch" it, and the electron shells are fuzzy, I don't quite see how we should tell that an electron belongs to Cl rather than Na. It's the same locations except at a few small positions.
But it does make more sense in BaTiO3 and SrTiO3, where Ba and Sr are not completely surrounded but have an eccentric position - or rather, two possible positions, which make these crystals piezoelectric, ferroelectric, and at the proper BaxSr1-x composition versus the temperature, gives them a huge permittivity. For these compounds, Ba and Sr do carry a (partial) charge, whose effect is observable by the strong polarization of the solid.
BaTiO3 and SrTiO3 being strongly nonlinear (they make the type II ceramic capacitors for electronics design) one could observe the saturation polarization, and from the atom movement, infer the partial charge. I stilll haven't done it and would be interested if someone did.
Same for less symmetric crystals like quartz which is piezoelectric. There we can tell that the bond is polarized.
And what about water, with its permittivity of 80? Its polarization results from molecule orientation, for which accurate models exist. Obviously H and O carry a partial charge. Should we call water an ionic compound because of this? No. It's better to say that bonds are more or less polarized.