The 'official' advertised long term rating of the speaker at 250W is clearly complete fantasy with a 1.5kHz tweeter X/over.
I understand that it certainly looks that way at first glance, but in this case I think the claimed power handling is realistic, as far as the tweeter is concerned. This is assuming the same yardstick ("music program", "continuous", "IEC whatever", "AES whatever") is being used for the tweeter's wattage rating as for the system's wattage rating.
[EDIT: My assumption about the same yardstick being used for the tweeter as for the system was
INCORRECT. I left the rest of the post up but with my initial assumption being incorrect, so is my conclusion.]
You see, there is a non-obvious factor which comes into play: The power distribution of music is nowhere near uniform across the spectrum. It goes down as we go up in frequency.
The ballpark rule of thumb I go by is that the wattage a speaker system can expect to see is equally divided at about 400 Hz, and is halved for every octave north of 400 Hz (it's more complicated south of 400 Hz. Anyway I've had less than 1% tweeter failures using this rule of thumb... but to be fair, I always err on the side of having at least some safety margin).
So if the crossover is at 400 Hz and the speaker is seeing 250 watts, roughly half of that will go north of 400 Hz and the other half south of 400 Hz.
The 1.5 kHz crossover of the Sointuva is approximately 2 octaves north of 400 Hz, so I'd expect the tweeter to see approximately 125/4 = 31.25 watts. (I'm using that 31.25 watt figure because it makes the math easy - as you will soon see, a few watts either way at this step won't matter.)
And now let's pad the tweeter down by 11 decibels, taking its 96 dB sensitivity down to 85 dB, and taking our 31.25 watts down to 2.5 watts.
The tweeter is rated at 80 watts thermal with a 2.5 kHz 2nd order filter, and we'd expect that to be halved (40 watts) one octave lower. If we only "need" 2.5 watts power handling for the tweeter, then the 40 watts we can reasonably expect translates to a 12 decibel(!) safety margin above the claimed 250 watts system power handling.
(It is theoretically possible for excursion to limit the power handling, but not in this case: The unusually large 3 mm peak-to-peak excursion of this 34mm diameter tweeter translates to almost 800 watts excursion-limited power handling at 1500 Hz.)
So I think March Audio is on safe ground with their claim... unless somebody hard-clips their amp.
[EDIT: The 11 dB of padding required to get down to 85 dB should increase the system power handling proportionally, which would in turn be reduced maybe 3 dB or so by the lower crossover frequency. So 250 watts system power handling still seems reasonable to me, but without the huge amount of headroom implied by my mistaken initial assumptions.]
(It can be scary to trust the math this much. I make a bass guitar cab rated at 500 watts whose tweeter is rated at only 20 watts. Using 200 Hz instead of 400 Hz as the "equal power dividing line" for electric bass, the math predicted the tweeter would have enough power handling but with little safety margin. Over the past twelve years I've sold hundreds of these cabs and only had ONE tweeter fail.)