Ok, it looks that the pin voltages you measured do make sense
Pin 16 is +3.3V in the MAX3232 package and it's ok
Remember not to connect it to the USB adapter: that's because the USB adapter it's self powered. There should be only 3 wires between board and adapter: TX, RX and GND.
TX and RX are most likely to be found among pin 12, 11 and 9
Pin 11 is the only (of the 3 pins) that is an input in the (not mounted) chip, so it is likely connected to the output of the processor. Here is where you will find if the processor responds to debug data, and here is where you can find the baud rate. it should also be connected to the RXD pin of the USB adapter.
pin 12 or pin 9 should be connected to TXD of the USB adapter, and start send random data throug hyperterminal, watching on the oscope if any response on pin 11. Which one? from the image you posted, following traces on the pcb, it looks that pin 9 is most probably the right one, but try also pin 12.
Now the bad thing is pin 15, since it should be ground, and placed @ 0v, while it's @ 3.3v
anyway, even this does make sense, since pulling it high might mean putting the (not mounted) chip in standby. The bad thing is that if the cpu puts it in standby it might not accept at all debug commands, since it believes that nothing should come from a shutdown chip
so if you always fail getting responses, than you need to find what triggers the debug port active. for example, try booting that device while pushing some key, something like "type del to enter the bios" on PCs. Of course, you should monitor pin 15, and when it falls @ 0v, it means that you performed the right action, and that the device is ready to accept debug commands
All I said might be wrong, because I don't have much hardware hacking experience, but that's my opinion.