As the capacitance of C1 is decreased, it does less and less as you might expect. And even without the capacitor, there is some top-octave compensation simply because the tweeter's voice coil is inductive and voltage division between the resistors and the voice coil tends to create a slightly positive slope in its response curve. But there is definitely more lift at 10kHz and 15kHz with a 0.47uF capacitor, and it is audible.I didn't look at my stock of cable assemblies this weekend, but I'd sure like to have a look at yours. It should be providing significant attenuation at 2.5kHz, and most of the attenuation should be removed at 15kHz. You can really hear the difference between such a cable assembly and a straight wire, so if that's not the case with yours, then something is wrong.
If you'd like, we can get together this week and have a look. Or we can check it out at Phil's meeting