So Steve81, you've heard the LS50's playing at full tilt? I know of no speaker with a 4-1/2" driver that could approach the lows that the LS50 reliably hits.
I've had several opportunities to hear them move some air; however, that's neither here nor there with respect to the facts/science of the situation. Anyone can go to KEFs website and check out a head on shot of the LS50 to confirm that the tweeter assembly takes away a significant fraction of the cone's radiating surface; by my reckoning using the power of MS Paint, it's about 25-30% of the diaphragm by area, or roughly the drop you'd expect in going from a 5.25" woofer to a 4.5" woofer. That the LS50 manages to dig fairly deep is a matter of T/S parameters and enclosure tuning, not Sd. Still, KEF can't cheat physics. Referring back to the FR:
By the graph posted at Stereophile, the LS50's port is providing its max output around ~50Hz. However, looking an octave higher to 100Hz, the woofer is basically doing all the heavy lifting. This higher range is quite important for several reasons.
1. It's usefully above the THX standard 80Hz sub XO, so a sub won't significantly reduce the burden unless you cross higher.
2. It's an extremely active region for all types of content.
3. More volume displacement is required to pull it off than most people realize.
So lets do a comparison using a basic
excursion calculator between a single 4.5" woofer and a pair of 5.25" woofers; to make things easy, we can assume both have 3mm of one-way excursion (note: this represents the Klippel tested xmax for the LS50's woofer, with the limiting factor being compliance variation). At 100Hz, at 1 meter, our 4.5" woofer can put out 95dB of sound, which is a far cry from the 106dB specified by KEF (with no qualifiers, making it worthless). Mind, 95dB at 1m is nothing to sneeze at, and in a smaller space, and/or for someone who doesn't need reference level output, it's probably adequate. However, that kind of output isn't going to cut it for everyone. In comparison, a pair of 5.25" woofers @ 100Hz @ 1m will deliver 103.5dB of output representing a gain of nearly 9dB, which is quite significant any way you slice it.