How can this be a crossover-less design and incorporate a tweeter bias function?
The panels are acoustically balanced to match each other, plus the audio transformer with it's RLC characteristic is making separation in frequency range in interaction with the capacitive load . Of course this separation between panels is not 24 or 18dB in octave. This separation is like first order crossover.The transformer has separate outputs for bass panels (higher voltage), and the tweeter panel has it's own output from the transformer.
Why would there be discussion of a tweeter adjustment if there is no crossover? How would one separate out the function of the tweeter for adjustment if there is no crossover?
The tweeter can be adjusted by its own supply voltage-(it has no effect on it's frquency characteristic) the sensitivity of the tweeter is adjustable without crossover of course. By this way you can have sound with more or less bass, because the sensitivity of bass panels stays the same.
The article includes this sentence, "The high-frequency panel measures 120x14cm and is designed to cover 200-20'000Hz," which seems to definitively show that it is not a crossover-less speaker.
As I mentioned it is a crossoverless design. The tweeter panel is acoustically designed to work from 200Hz to 20KHz.
Even if not crossover-less, it does have a couple features not normally found on ESL speakers. The impedance matching seems to be a significant feature.
The impedance of Virtuoso is like a GAUSS curve (0 Ohm in 0Hz). It is an electrical proof of the crossoverless design. If you have some crossover in the circuit, even one capacitor, then the impedance will increase by going down in frequency (in bass region) See some full range ESL like Soundlab, QUAD or others. It is not the case with Virtuoso.