Yes couldn't really work out what the 15L104W is there for. It isn't connected to anything other than the serial ports on the 328. The only thing i've seen it do, with the original firmware, still on the original 328 but not readable, is receive some serial input from the 328 and then respond, happens every time its powered up. The 328 send whats looks like some binary data that finishes with the ascii text "ending." the 15l104 replys with an ascii number and "OK" in ascii from memory something like "3624OK!". It didn't come with the header pins BTW I soldered them on.
Re the spi I kept getting a blank white screen trying the hardware spi when bitbang displayed stuff. I saw a suggestion I think you gave someone else in here to try reducing clock rate so I did quarter worked and half worked. Possible if I go back to default it would work now because I solved another issue with my confusion over the CS and Reset pins after. But working fine as is.
Oh I also changed the 5v reg, supply not really high enough for the original, for a better LDO one and adjusted the UREF_VCC to reflect the actual when measured, not sure if should do that? Seems reasonably accurate compared to other measure kit have access to.
Also nearly forgot to mention the IR receiver function didn't work at all with the orignal firmware, no sensible response from supplier. I could see sensor OK with scope. Part of the motivation for the firmware change, it works fine now.
Cheers