The problems start at high frequency in the I/V converters. Over 350 ohm you have a big capacitor if you want a low-pass filter. That capacitor is between input and output. The output stage must drive that capacitor. After that, at the output of OPA1612, there is onother low-pass filter, R=100 ohm + C=10nF if I remember correctly. It is not easy task for the opamp.
The problems start at high frequency in the I/V converters. Over 350 ohm you have a big capacitor if you want a low-pass filter. That capacitor is between input and output. The output stage must drive that capacitor. After that, at the output of OPA1612, there is onother low-pass filter, R=100 ohm + C=10nF if I remember correctly. It is not easy task for the opamp.
Now we have a better communication basis. Because of the lack of the schematic, we can only make some assumptions about the reality. Anyhow, if I correct interpret your post you see the problem of this type of DAC's in the output drive capability.
As I know, Amir is measuring balanced outputs with around 2x 47k. This was (is) since a long time the standard input impedance of amp's. So we will see no problems with the measurements. Because a wide range (all) of DAC's will not be really loaded with this impedance.
To the feedback capacitor, we can see different values from manufacturer to manufacturer because it is also layout depending. As well, we can see that the Ess DAC's have a problem with the output stage (Hump) and from this case, the 1612 was the best choice for I/V convertors (bandwidth…)
The output impedance is for my application from interest because I would like to run a Purifi amp in direct mode with its much lower input impedance. I guess (hope)with 4 ops parallel, it would be a manageable load for it.