@Tachikoma
I actually ended up returning the CRBNs due to the very minor distortion on a few tracks, and ordered some Stax headphones (can't wait for them to show up).
In the meantime, I've done some testing of the iFi Pro iESL based on your thoughtful post. Measurements were done using two AHB2's, each in bridged/mono mode.
I do indeed measure about a 10mV DC offset if I really crank the volume to uncomfortably loud levels while using the iESL to drive some Sennheiser HD800's. For almost too loud measurements, they range from 5mV-10mV (no clipping at all from the amps though as I'm running two of them now).
Further, with or without the HD800s attached to the iESL dynamic 4 pin XLR on the front, I get similar measurements for DC offset that fluctuate based on the music between 5mV - 10mV with the HPA4. Thus, once I found this out, I decided to do some clipping testing with no headphones attached to the iESL as I didn't want to risk damaging my headphones (e.g. amps would then see only the input impedance of the iESL as I left the iESL turned on). As you suspect, clipping does occur near or above 0dB on the HPA4 (a level that would be unbearably loud with headphones on or even just my speakers).
When I measure the DC input impedance at the iESL using my multimeter (amps are unhooked from the iESL), I see 11.91 ohms for the 16 ohm setting (where I usually keep the impedance selection dial on the iESL); 25.23 ohms for 24 ohm setting; 26.22 ohm for the 64 ohm setting; and 39.55 ohms for 96 ohm setting. These measurements were without any headphones attached. If I attach headphones, and redo the input impedance measurements, the measurements are slightly smaller, ranging from 11.83 ohms (16 ohm setting) to 39.24 ohms (highest setting of 96 ohms).
That said, I assume your post meant putting a resistor in parallel with the load the amp sees (e.g. in parallel with the input impedance of the iESL). I do have some 4.5 ohm 12 watt resistors, but don't have anything of higher wattage. Assuming 4.5 ohms, this would lower the impedance the amp sees to around 3.26 ohms if I keep the iESL impedance selector on 16 ohms (4.5 ohms in parallel with 11.91 ohms).
I'm hesitant to do this, unless I ever need more volume out of the current set up (not sure what headphones this would ever be with two AHB2's). If I ever go back to only one AHB2, it definitely seems a large 4-6 ohm resistor in parallel with each speaker input on the iESL would fix the low volume issue when driving some headphones like the CRBN (one AHB2 amp is actually rated to drive a 3 ohm load at 240 watts:
https://benchmarkmedia.com/products/benchmark-ahb2-power-amplifier).
I am curious on your thoughts...
I think as long as there is no clipping at loud volumes (e.g. above normal/loudness is still bearable, around -10dB on the HPA4 volume), my meter shows less than 10mV DC offset with a track playing at the loud level, and no distortion is heard, I am OK unless I go back to one amp... In the current setup, it seems the clipping lights on the AHB2's will alert me to whether the iESL sees any large DC offset.
I haven't taken the iESL apart yet, but it does seem the impedance selector's impedance would protect the iESL's transformers regardless?
Thanks!