My experience of the Radio side of the codes / tech2 may differ from these displays, but what I know so far. . . .
One of the other models in our research dept at the moment, The DVD90, appears to effectively load an eeprom on power up from the flash and Tech2 reads and writes to the eeprom, not the flash, but when powered off, the eeprom is emptied. The flash has multiple data lines in to allow for programming, but it is difficult to access all areas without all the data lines. The PIN code is retained inside the part of the flash memory with data lines fused. We are using a secondary chip piggybacked on the top to monitor data in and out, and its by reading this afterwards that we can get the code. A quick calculation from an algorythm we have here then gives us the PIN code from that data.
Exceptionally long winded, and so far the hardest radio code I have ever had to crack - if the CID displays are similar, then it may be a case of reading them "live" rather than chip off.