I suspect that he engaged the rolloff filter, not turning off the tweeter. I've run guitars through the AF/TC with Bill Lawrence's new microcoils through low capacitance cables and gotten outstanding results.
I'm also strongly considering getting another AF/TC to use as a stereo setup as I like programmed music through it more than through my KRK studio monitors. We'll see what space and money I have when I move into a new apartment in a month or so. I feel like I'd be moving from a sweet spot a couple inches wide to one a couple feet wide
At age 57 it's fun to still have new and enlightening listening experiences. Over the years I've heard electrostatic speakers many times. A friend's ancient Acoustat 3 panels sounded very good. Another acquaintance about 20 years ago had a pair of huge Sound Labs with electrostatic sub woofers, but it sounded pretty bad, likely needing a lot of setup help.
Duke's respect for stats rubbed off on me. They sounded excellent at Duke's place, and even better at one CES setup by the late Bob Crump, co-designer of the Blowtorch preamp with John Curl (legendary high-end audio names). Ray Kimber's 4-ch system with two huge Sound Lab stats in each corner playing proprietary "Iso-Mic" Kimber recordings is a high lite of my listening experience (that system will never again be exhibited, about $330k value).
As hard as audiophiles will resist this statement, there is a lot of the mid-range performance of the best stats in Duke's current technology including and especially the AF/TC. I used to think this mid-range characteristic of the stats (for which they are legendary) was from the light weight of the diaphragm, resulting in extreme speed.
Duke's 12" woofer cones are obviously
many multiples heavier than a stat panel, yet Duke's system provides a very close replica of the stat's mid-range performance. So I now presume low-weight diaphragms are not a prerequisite to achieve stat-like mid-range performance. Rather, the mid-range performance seems related to dynamic power and a particular, linear radiation pattern (obviously a stat employs a sum total one diaphragm and has no Xo splitting the signal, with all it's phase/radiation pattern artifacts, etc).
In the bass range TC comprises a juggling act including sensitivity, power handling, cabinet size/weight, and bass cutoff rate/frequency (users may fine tune these parameters by plugging one or both ports). Duke's further unique expertise was to translate decades of high-end audio fanaticism to the mid-range/treble performance and Xo range between the two drivers.
I apologize if all this was covered earlier. Persons familiar with Duke's high-end audio speakers will not be too surprised at TC's unique excellence in the music/pro sound field. I estimate he spent a good year considering the bass parameters mentioned above, then a good year longer achieving his mid-range goals, which were obviously much higher than any prior music-pro sound standards.