Pass labs xa25 amp and BW 804 d3?

As I said, I know Pass overbuilds his amps, and you might not have any issues, but once again, why tempt fate?
If it were me, I'd take the safe route and get a higher powered amp. I believe in having lots of headroom when it comes to power.
 Then if you decide to change speakers,you'll have a wider range to choose from.
I see some XA-30's for sale that are under $5K, still low, but closer to recommended specs. There are a pair of XA-100's, asking $6.5K though, but would be a real killer.
There are other manufacturers, but I don't know if Pass is the only one you'll consider.
Bob

Gweedoargus,
You have gotten some good responses and comments. I’m inclined to agree with those who have raised the very important consideration of ’sound quality ’ as compared to increased quantity (higher oower/more watts). The Pass Labs XA 25 is capable of 50 watts of pure class A into a 4 ohm load (this isn’t an insignificant amount). It can provide class AB 200 watts maximum into a 2 ohm load (This is impressive).

The John Curl designed Parasound A21 offers class AB 250 watts into 8 ohms and 400 watts into 4 ohms. So clearly higher power. I suspect it has pretty decent sound quality. Keep in my its DF (Damping factor) is >1100!! This suggests very heavy use of NFB. Some very respected amplifier builders such as Atmaspere and Roger Modjeski consider this amount of DF gratuitous and without meaningful benefit. On the other hand some builders (John Curl I’m sure) would beg to differ. My gut feeling is that the Pass XA 25 could very well be the superior sounding amplifier.

Is there the interest/desire to home audition these amplifiers and actually hear them driving your B&W speakers? I would bet that the vast majority of your listening the XAX25 would remain in class A operation. This is a good thing IMHO. I'm expressing my own biases. I don't believe in the cliche of "the more power the better" once you have sufficient power the objective should be placed on quality. 
Good luck,
Charles

I consider that kind of damping factor pretty ludicrous too. 500 is pretty typical for Pass's mainstream amps, and the XA25 certainly accomplishes that in no small part to lacking any kind of degeneration on the output transistors. 

If the XA25 isn't going to make the power you need, the XA 30.8 isn't either. But that's not why you should choose between them. They're totally different amps. The XA25 is a conventional push-pull topology with negative current feedback. The XA 30.8 is a balanced topology using Super Symmetry feedback. They work and sound totally different. 

I just crunched the numbers for those speakers. 25 watts will yield 101.2dB at 8 feet. That is insanely loud. Nobody listens at that kind of average volume. And 25 watts is just half of it's class A envelope. I crunched the numbers on the amp, too. It seems to be running 24V rails and hits 1% distortion into 8 ohm at about 22.8V. It's current limited to 10 amp, but you'll never see that with those speakers. At their 3 ohm low point that amp will drive 173 watts into them with current to spare. That's like 104dB at 8 feet. If you typically listen at an average volume of 92dB, which is still pretty loud, you only need THREE watts with those speakers at 8 feet. 

Kosst_amojan,
I agree with your calculations and reasoning. This is the point I was attempting to illustrate in my prior post. You can’t look at the XA 25 and conclude, "oh, it’s just a 25 watt amplifier " . There’s considerably more going on here if one can appreciate the design/circuit and implementation. As you calculated, at very reasonable (even generous) volume levels these speakers aren’t demanding much power/ watts at all. And as a reminder, 200 watts/10 amps  (current max) into a 2 ohm load.

DF of 500 is still pretty high relatively speaking in regard to the XA 25. I agree the XA 25 and XA 30.8 are notably different circuit designs. I’ll empathize again, Parasound A 21 or one of the Pass Labs amplifiers, you have to hear them with the specific speaker of interest. Each amplifier will have its own sonic signature with a given speaker.
One point to consider regarding DF is it is inversely proportional to the output impedance (isn’t necessarily a bad thing) and the higher the DF the more NFB utilized (Which ’could’’ introduce some undesired sonic consequences of its own). Another point,  speaker sensitivity measurements are of a single speaker.  Generally 3 db is added when discussing a "pair" of speakers. 
Charles

Seems like Pass has some very knowledgeable owners. 
If they say it can work, who am I to disagree?
B