As stated before, the output has not fallen - just the input!! The fact that they have identical sound pressures would suggest that the actual power output to the speakers is pretty similar!

As far as power is concerned, expecting the BL4000s to go very loud in a large room would be unrealistic but for the task for which they are designed, they are fine. Remember that most music when listened to will use less than 1W. The bit that really uses the juice is however, as has been pointed out, the bottom end and it is of course perfectly feasible to add a BL2 to your 4000s to give some real power! 

Anyone has a feeling that they will upgrade the Beolab 6000 to be better sounding than the Beolab 4000? I simply admire the design of the Beolab 6000 but not very convinced with the sound. 

That's always going to be difficult. Slimmer cabinets means smaller drive cones, which means less cone excursion and higher resonating frequencies. It's not all about bass though, and I suppose smaller cones also means less cone breakup and hence better midrange. Can't be sure if they will update the 6000s, but it was mentioned a while back on the forum...

Beolab 6000 has two and not one mid/bass drivers so cone exursion will be distributed leading to lower distortion because of the greater sum of surface area. They can try replacing those cheap drivers though. I wonder why hasn't B&O pass all the manufacturing of drivers to scanspeak yet since they have a long history of using their drivers, B&O can certainly take advantage of their technologies and not necessarily raising the costs of production.

Fact is that B&O use very cheap drive units in these 'lifestyle speakers' I think for B&O to be taken seriously they want to stop the 6000. It is not possible to produce the standard required with CHEAP paper drive units that measure 3 or 4 in.