Since we don't know how much of Flash ROM is used were guessing at the amount of extra functionality that can be added. There seems to be enough ram (32MB) to deal with data even an FFT. DSP Bandwidth wise I would say that there is sufficient and can always be extended by increasing the refresh cycle currently set at 1.5 msec. The unpopulated circuit I'm guessing would have been a 12v supply for Firewire. There are actually 10 diodes 1 for power and two for Ethernet that are not shown on the pic (small PCB extension with connectors). Couldn't power unit up while apart due to short power cable. Sorry all put together again. I do have higher res pics but wanted to save space on Dropbox. Ironically the USB cable from the paddle board to the mobo is shielded. Why not continue the shield to the iPad? I was tempted to mod the unit but it's going on the chopping block. I don't have the charging voltage values that are on the USB Data lines to see which charging current is selected by the DL but it appears that it may be too high thus causing excess noise. This was introduced by the recent iPad SW update where they finally corrected the charging current sensing problem. I doubt that the DL even looks at the voltage on the USB Data lines for the correct amount of charge current requested it just tries to supply it and the filters in place may be inadequate as you stated. A couple of well placed ferrites may help.
