It drives condenser mics well.
Which is not a major challenge per se. Are the opamps next to the mic inputs 4580s as well? Should be fine for condensers, but don't expect much in the dynamic mic department.
As for line level inputs, well, I have worked out that it doesn't blow your gear up but that 48v causes lots of distortion. I ripped out the 48v power supply and am in the process of rejigging (butchering) this to 5v as that suits my current needs best.
Lots of distortion? That's a bit odd, but I guess it would depend on the output. P48V short-circuit current is only about 4 mA, and I would expect the line-in portion to have several 10 kOhms in series so it would be a fair bit lower there (maybe around 1 mA). Before ripping out the phantom power, I would rather make some sort of external adapter with some maybe 100k pull-down resistors and about 22 µF 63-100 V coupling capacitors, + facing the input. There would be enough space for some coupling capacitors inside the unit by the looks of it, assuming you find a good place to tap into the line-in signal path (maybe at series resistors).
Interesting design for sure. I wouldn't be certain that the unit even has an analog input gain control, though a gain pot could be running over the edge connector of course. If it doesn't, they're using the dynamic range of the AK5552 ADC to make up for it (would that even work?). The output side must be using the DAC built into the ADAU1701, with U11 being the headphone driver. I'm not sure what exactly the MVSilicon BP1048 Bluetooth audio processor (DSP) is used for, which also seems to sport a DAC of its own though it's even more basic than the ADAU's.
It's a surprisingly barren-looking circuit board for an audio interface in any case. Some of the magic may be hiding on the secondary board that houses the encoders and stuff.