1. Wharfedale Opus 2-M2 with Legacy Audio XTREME XD Subs
2. King Sound King III electrostatic with same subs at same settings
3. Ohm Walsh Model F with same subs, same settings
All systems used same source, Small Green Computer sonicTransporter with SONORE Signature Rendu with systemOptique; same USB, Clarity Cable Supernatural 1m; same DAC, Exogal Comet with PLUS Power Supply, same amps, Legacy Audio i.V4 Ultra, and same cables, Iconoclast Cables and BAV Power Cords.
Ideally the sub would be dialed in for each, obviously, but my latest interest is hearing each for the imaging and soundstage characteristics pursuant to this discussion. If you wish to argue with me about the method employed, I'm not interested. Build your own systems and do your own direct comparisons to derive your conclusions. :)
Wharfedale setup: Quite lovely with the expected density of center image due to dynamic and 3" mid. Even with a very generously appointed midrange and bass driver, it becomes evident that bookshelf speakers are limited by their small cabinets. The driver selection is constrained and the performance level dragged toward the perhaps upper middle of the performance spectrum of HiFi versus the practically limitless drivers and designs of larger towers. There are very good reasons why most manufacturers do not opt for bookshelf systems as their ultimate expression of speaker building.
King III: Instantly obvious grandiose soundstage which makes the bookshelf's sound stage seem diminished, and with much superior spatial location of performers. The depth of soundstage, the vastness of venue, becomes quite evident. Center image is stretched, obviously, due to the driver type and configuration, however more finesse, fine detail is heard in the center image. Far more consistency across the acoustic plane from L/C/R, versus the hot spot of bookshelf imaging.
Ohm Walsh Model F: It can be surprising to hear how muddled this driver is in comparison to the others. The promotion of these older drivers as having exquisite resolution in comparison to newer technology is simply wrong. Quite a muted mid relative to the others. It becomes obvious that this older driver has serious constraints. The upper end is clean, light, but the midrange on down is hindered, even when carefully double wiring it, as with the others. Here is an interesting point; the vaunted omni sound stage with this speaker is not all that; the King III throws a significantly larger, more voluminous sound stage and with much better cleanness top to bottom. The ohm character is the antithesis of the Wharfedale's immediacy, and the King III masters them both. I would not choose the Model F as a reference speaker. It's just a fun play around speaker for me. The King Sound King Tower omni has superior openness and cleanness.
One outcome such comparisons is that the audiophile should not think to recreate the performance of a tower by swapping in bookshelf and sub even with a larger bookshelf speaker. Not going to happen unless the largely unproductive goal of recreating the precise set of drivers and crossovers of a tower speaker in two cabinets is pursued. At that point you have largely failed at advancing the system by splintering it. Genre of speaker is crucial to outcome and imo should color the discussion of tower vs. bookhself/sub as much as anything.
There has been a fair bit of conversation about subs with tower speakers. Big towers do not necessarily need subs, but in terms of macrodynamics they almost always are a boon. However, you will pay a price for slapping additional big drivers into the rig. Coherence will suffer with subs added (No, I do not care about the opinion of those who flog the topic of distributed arrays; they have their own idiosyncrasies, and I am not interested in debating my discussion. BTW, I like Duke Lejune and his idea, and have heard the DBA prior, but see pro/con in all sub setups), so it's a trade off between presence and precision. Imo, not possible to get perfection on both spectrums simultaneously, and trust me, I've tried dozens of times with systems.
I offer this as an example of what might be expected in direct comparisons between different genres of speaker systems, especially in terms of center image and sound stage. Obviously, particular characteristics will change based on selection of genres of speakers, but I would not expect the framework of the outcome to change.