Magnepan 1.6 active 24db slop better than stock?

The electrical crossover is 1600 (maybe 1500).

The accoustical crossover is where the woofer and tweeter SPLs are equal. If you have a spectrum analyser, connect the tweeter with inverted polarity and look for a deep notch in the frequency response.

The inverted polarity of the tweeter is correct for the stock tweeter crossover (12 dB). For a L/R 24 dB crossover normal polarity is correct. (Otherwise you get that deep notch).

Stereophile test shows x-over freq to be 600hz.
Knee frequencies are +- about 100hz from that.
Hi pass is 6db oct consisting of caps ONLY
Lo pass? 12db per octave and includes an inductor.

http://www.stereophile.com/content/magnepan-magneplanar-mg16qr-loudspeaker-measurements

As kind of an aside, I think this accounts for the tweeter sounding better on the inside, being 90degrees out of phase with the bass part of the driver. The tweeter being a couple inches closer to the listener, straightens that out.
Also, please note that this is NOT a difficult load. Very flat impedance with only a peak at crossover while phase angles remain moderate. Low sensitivity? Sure, but this is a line source / dipole and will react to a room somewhat differently than a box with forward facing drivers of similar electrical measure.

Best of luck with your experiment. Please post some results when available.

Magfan... Yeah. I got ir backwards. It is the woofer that is 12 dB slope, and the tweeter 6 dB.

As to the crossover frequency, 1600 is what you get from calculations using the values of components in the crossover. The accoustic crossover includes mechanical characteristics of the drivers. No doubt that the accoustic frequency is what matters when listening to the stock speaker, but for designing a new crossover the electrical crossover is important.

By the way, I am a bit leary of driving any tweeter directly from a power amp. Power amps are apt to output nasty thumps.

With 'ya on the thumps. A nasty thing to do to a tweeter. How about a turn-on delay relay? That help?

As for the x-over. I'm going with the measured results from Stereophile. The 2 drivers have very equal output at about 600 hz +-20hz. They are individually several db below 'average' output, and so the sum is pretty flat thru the passband. (right term?)

Also, consulting the measured data, it makes me even more convinced that changing the inductor to an ultra low DCR value is a mistake, raising output below crossover. This is already shelved UP vs the above-crossover frequencies.
Stock inductor is 16ga, iron core, of 0.40ohms DCR. You can easily buy an inductor in various configurations, down to <0.20 ohms DCR. No thought is apparently given to coil shape optimization.
My simulation yields a 14ga coil with as many turns as layers.....theoretically near-perfect. And, it is aircore with a DCR of about 0.38ohms, and will be inaudibly louder than stock. But, with near-zero chance of saturation at the very highest levels. The OTHER advantage of my coil, is that it uses Standard values of PVC pipe for the core.

Magfan... I redid my crossovers using an aircore inductor with lower DCR (don't remember the value). However, if it slightly boosts the woofer output this means that there is less need for a tweeter pading resistor. Overall results are good. The speaker measures flat, but by the time that room equalization is done flat response is not needed.