(the only 15" P-Audio I have are SN15C) - for a minimalist cap-highpass-only situation, a swamping resistor may help the HF smoothness part and reduce some unfavorable Z-interaction. A measurement system could help you with the woofer lowpass selection.
Which highs are you referring to? The ones over 10k or the ones below?
If you need to level down the ones over 10k then a pad will be a good idea.
Otherwise, just decrease the value of the capacitor.
Anyway, i don't think that filtering those speakers with a simple cap will work.
At my current version i am using for my bm12cxa second order on the woofer and third order on the driver. I tried using first order (single coil and single cap) and i don't like the way they sound. The driver distorts at high power.
I also don't like the way the driver(bmd440s) sounds if it is playing at 1.5kHz. It sounds like something it's ringing. If you look at the impedance plot you see a peak there, so it;s probably a resonance. It's also showing on the CSD plot if you measure it. Crossing the driver with a third order crossover solved this problem (for me).
If you are using a first order crossover it's likely that you are facing the same problem.
And of course massive overlap between the two speakers with lobbing and comb-filtering.
The problem here is that US seller Common Sense Audio recommend a simple 1 cap crossover for these. While it does "work", it does not sound good at all.
This has lead many builder astray. The P.Audio coax are being promoted as simple to use with only a cap on the tweeter. This does a huge disservice to these very nice drivers. A simple 5 or 6 part crossover will be worlds better than the single cap. The drivers deserve it.
IMO it would be great if Common Sense Audio would develop (or pay to have it done) some good passive crossovers for these drivers. They could even sell kits. It would be good for everyone.