McIntosh and Autoformers. . .?

According to what I have read in the past, an autoformer is either a straightforward output transformer - albeit one with bifilar winding and multiple taps - or it's different in that it doesn't have a secondary coil, which makes it sound like an iron-cored inductor of sorts. Regardless, any type of output transformer is a rare thing for a solid-state amp, and the output of Macs so equipped is fully balanced rather than hot-and-ground. The point of having them at all, as far as I've been able to glean, is to present the amplifier output stage with a constant impedance to work with, optimizing linearity and power transfer, and also to block any potential DC at the outputs without having to resort to capacitor- or servo-coupling. Mac states that their autoformer models are fully balanced amplifiers, with the phase and anti-phase signal from either of the two duplicate amplification paths being combined at the output autoformer's bifilar windings in order to cancel noise and distortion common to both legs. Though Mac says the autoformer has wider bandwidth than the amplifier itself, audiophiles have theorized that its similarity to the output transformers used on typical tube amps gives these Mac models a superficially similar sound, compared to typical SS amps where the ouput stage is more directly coupled to the speakers. I hope someone with the technical knowledge comes along to shed a little more light on this question.

Thanks all for offering clarifications about the DC-blocking capabilities of the autoformer - I was wondering how this could be so in the absence of a secondary coil. Nealhood, just as a side note, I believe Mac still claims 35dB of negative feedback with their AT designs. Also, why should we consider AT's "mandatory" for OTL's, provided the partnering speaker presents an >8 Ohm average impedance?