Okay. I had better luck today searching the internet. Here is a PDF of the chip itself.
http://www.holtek.com.tw/pdf/computer/82m98av200.pdfYou can see in the reference schematics that the 6 MHz crystal is tied to pins 17 & 18. If you have an oscilloscope, you could verify that it is working. I doubt that it is bad as you stated that the buttons work. That means that the chip is working in that respect. If the buttons are working, then the volage at pins 1 & 16 are correct, there is a 6 MHz signal at pins 17 & 18, pins 12, 13 & 14 are receiving the porper voltage during their respective button press and pins 3 & 4 are sending the correct data to the USB post on the computer. This rules out the chip for the most part.
This leads me to believe that one of the optocouplers IS bad as they are tied together. I would imaging just by looking at the pictures that if the optocoupler on the first board is bad, then the second one would not work.
It looks to me just judging from the photos that the LED in the optocoupler on the main board is closest to the chip. If you place the black lead of your multimeter on the black ground cable of the plug and then probe with your red test lead (using the DC setting), you should see 5 volts on the red wire. If you follow the trace to the optocoupler, you'll run into "R1". You should see 5 volts on one side of R1 and a different, smaller voltage on the other side. If it's the same voltage, then something is open. Just keep following that trace until you see "0" (zero) volts. Once it shows zero, then you know where the open is. It's probably going to be one or the other. Just don't remove the black test lead from the black wire and you should be fine.
Let me know what you find out.