I have described clicks occurring occasionally during changes of the filter types for bands 1 and 5 if a signal is present via mail to MC.
With the new firmware, I can still reproduce it.
Using -20 dBFS RMS pink noise, this even produced a click that overloaded the device. Output was Line Out, Vol was -32 dBr, PEQ was enabled and Loudness was enabled with boosts at default values (7 dB) and Low Vol Ref at -38 dB.

Even without any EQ of any kind except band 5 at 0 dB gain at 5 kHz with Q of 0.5, it produced a click indicated at ~12 dB above the level of attenuation, ~-9 dB for Phones at -36 dBr.

The parameters for the PEQ for Line Out are:
B       G         F      Q
1   -2.0      150   0.5
2   -2.0      150   1.3
3    0.0      500   1.0
4    0.0   2.000   1.0
5    0.0   5.000   0.5

All tests were done at 44.1 kHz sample rate using USB input.

EDIT: Added that Clicks appear with signal present. Edited part is in italic.
EDIT2: I don't think this would significantly change MC's response. So I will address it as is.
EDIT3: Clarified that I also changed filter type with Phones although all filters were set to 0 dB gain. Again, edited parts are in italic. This omission must have put MC off.

I see that I should have written more clearly that I also changed the filter type with Phones although all filters were set to 0 dB gain. Technically, in its original form as quoted by MC, it is incorrect, as I did no such test.

I have edited the previous post accordingly and hope that this clears up the confusion that may have arisen from my poor writing.

I sent you four mails in said exchange. I sent these mails to the mail-address which you used to write me in the first place and also today. All the mails I addressed to you using that address are marked as sent on my end. 

I will forward you the message (No. 4) that you must have missed. It says less about the clicks from the PEQ filter type change than my previous post but also includes some feedback on USB Firmware revision 25 (which might be relevant for USB FW 26).

I suggest that if we want to determine what might have went wrong with the mail exchange, that we do so in private.

Occasionally, switching the PEQ off also produces a click.