The tweeter does roll off naturally starting at 1k and looks like 3db down around 750 pHz.
The cap is 33uf and impedance 8 ohms. That calcs out to 600hz 3db down.
So it looks like the cap is slightly below the natural roof off.
I still thinks this I quite a stretch of definitions.
The two other drivers are 15inch each has a different slope
One rolls off at around 200 the rolls off at 600hz

Either someone is lying to you, or it's not merely a capacitor.

...But there is something that makes one focus on the tweeter while music is playing.
What I heard at CAF and at Florida was magical. The speakers just disappeared.
Not happening here.

Roscoe,

That's my understanding of the system. A capacitor wired in series with a purely capacitive tweeter. Dhar and I measured one at his house last winter and it appeared to be accurate. We measured with a 5uf cap then a 4 and the frequency response was identical but 1.5db down on the 4. They overlayed perfectly. Shrug shoulder emoji

Actually I have been working with Walt on this tweeter.
My big problem with the speaker even after almost 200 hrs of break-in is the tweeter is still calling attention to itself.
We looked a lots of things and added a 400 ohm resistor in parallel to reduce electrical resonances. We have added second order low pass to roll of above 22k the nasties with cone break up at the top.
These have helped quite a bit. But I will remove them and listen as a reality check to see if the changes I heard are due to mods or break in.
Since this driver is a 1.3 inch hard dome designed for handling the freq range of 575Hz on up, i am curious how much effect this non crossover 33 if metalized poly cap has on the musicality This is 90% of the sound.