Originally Posted by Bill Fitzmaurice
If said towers and bookshelves cost the same then the bookshelves should work better in the midbass and higher than the tower, as the book shelf components should be higher quality. There's no guarantee of that, of course.
But the devil is always in the details.
Let's imagine a reasonably high-end set of speakers: which is to say a set with rather expensive drivers.
Now: let's compare a 2-way bookshelf that crosses over a ribbon tweeter to a 7" woofer to a transmission-line tower that crosses over the same ribbon tweeter to a 5" woofer.
The 5" driver clearly has advantages at high frequency over the 7", meaning that we can accomplish some combination of better midrange (by using the 5" rather than the 7") and better midrange/HF (by crossing the tweeter higher). Now the cost of this is that we need more wood: but if we assume the costs of the speakers are driver-heavy (they don't use expensive laminates or the like), that could be entirely offset by the driver change.
Again, let's assume a reasonably costly speaker setup. In order to push the crossover point down on their 2-way bookshelf (let's say, using a ribbon tweeter and a 5" woofer), the company puts in two woofers. On their tower, they again create a transmission line, allowing them to revert to a single 5" woofer for the same F3. Since the drivers are reasonably expensive, the extra cost in wood is more than offset by the savings of a driver.
A company attempts to build an inexpensive 2-way bookshelf. Because they need to accomplish their range (say 60Hz-20kHz) in two drivers, and in order to keep quality, they invest in some rather expensive drivers. In designing their towers, they have room to go with a 3-way design. As you've mentioned, they use cheaper parts; but because the individual drivers each now have lesser needs, the less expensive drivers perform as well or better.
Building a 2-way bookshelf with limited space, the company chooses to use a single tweeter and a pair of 4.5" drivers (in order to get authoritatively to 80Hz). When they build their tower, there's space to setup a larger driver (say an 8"), as well as space to setup a TL design. This allows them to reduce to a single 4.5" driver (having offloaded the low frequencies), and perhaps even reduce the cost of that driver as they no longer needs one which performs as well at LF.
As I've said before: the implementation is more important than the theory