One thing to keep in mind with any add-on supertweeter that was not designed as part of a speaker system is that the transition between the main speaker and the super-tweeter is going to be widely variable.
The Townshend, for example, says its range is 20KHz and up with a -3dB point at 12KHz.
The Tannoy (depends on model) has a third order crossover starting at 14KHz. That means it'll be 18 dB down at 7KHz but there is still a lot of energy output below 14KHz as it comes up to crossover frequency.
In both cases, the tweeters have significant output in the range where many main speakers already have significantly flat output by themselves.
If coupled with a speaker with significant high frequency energy, you would have a rising frequency response above 10 or 12KHz. If coupled with a speaker with dropoff in highest octave, you couple have a dip in the high response with the frequency response then rising above the dip.
What is simply a roll of the dice is whether your resulting high frequency response bears any resemblance at all to a uniform, smooth output. Most serious speaker companies put a tremendous amount of effort into getting a seamless transition between drivers. With an add-on tweeter, that simply isn't going to happen unless by sheer accident.
While the idea of extending a speaker's high frequencies is admirable in theory, this seems to be a bit of "throw a spit wad and see if it sticks" approach. I find it interesting that basic preamp tone controls have long since been banished from high end equipment but audiophiles have found all manner of new ways to tinker with tone balance through products such as this and others.
The net result is that you can only try a unit and see if you like it. Just don't get the idea that you are necessarily improving the "accuracy" of your high end response. ;-)
