Absolutely not! Class D would destroy the transient timing accuracy, making the WTA filters almost worthless.
Their is only (for SE) one amplifier in the DAC, which combines the DAC current to voltage (I to V), filtering and output drive into one amplifier with a single global feedback path - and this path only has two resistors and two capacitors. This simplicity means that the output is extremely transparent.
Even though the global feedback path is short and direct, the amplifier itself is quite complex with a second order analogue noise shaper topology. Conventional amplifiers have poor feedback at 20kHz, and so it suffers with increasing distortion at high frequencies. But my second order noise shaping approach does not suffer with this, as the open loop gain at 20kHz is 120dB - that's a 20GHz gain bandwidth product for the technically minded - this means that there is no increase in distortion at high frequencies.
This is a particular problem with conventional power amps - as current is drawn from the output, the output stage switches from class A to class B, and then creates crossover distortion - which is very nasty, as it can extend to infinite frequencies, and can cause timing problems too due to delays changing with signal level.
Talking of output stages, the OP stage in TT2, like all my designs, use discrete transistors. The transistors used are very high speed (30nS) high power (10A) devices, with 4 devices used per output.
You can see the measured benefits of this approach from the slide below:
In this case the balanced outputs is used with 12W into 8 ohms. You can see no noise floor modulation, and absolutely no distortion above 3 kHz, proving that there is no crossover distortion at all. Indeed, when you remove the 8 ohm load, nothing changes at all - except for a small improvement in second harmonic distortion.
Eliminating noise floor modulation, and high frequency distortion in a DAC is unique - let alone one that can drive significant power into 8 ohms.
So no it's certainly not Class D...
