and verify the effect "in your circuit, with your drivers". I was playing with some Altec 902's, with non-standard 35480 diaphrams.
They roll off a little sooner than my favored 34647 diaphrams, but are a little flatter across the main operating range. Now comped with 1.0uF, the meter is more or less "flat" in room at 12 feet when flipping from 100 to 1000 to 10,000Hz. (-2db). At above 15kHz, the HF driver is on it's own response wise, since the only "boost" available is already in use by then...well, you know, the 10dB lost in the series R has been "reclaimed". Honestly, I think Wayne's PSD2002 comp works FINE with 0.33 or 0.47uF and 12dB of pad in my quasi-4 Pi Pro's. IIRC, the 0.47uF Dayton I installed across the 23.5ohm series R (a pair of 47ohm Mills) may even have been a little "too hot" at the upper end when measuring the system response.
In retrospect, my Altec's using the 35480 diaphrams have been slightly "hf under-comped" all along, making the 4 Pi's PSD2002's sound "brighter by contrast". Well, they REALLY were playing louder up high (15kHz) than the Altecs after all:) You didn't hear that from me, shhhh. Sam