I'm rather puzzled by this. The PIC is the same one, running the same code, as the original empeg tuners. I programmed up a batch of about 200 some years ago, and have been using that stock ever since. I know for a fact that several people who bought the kit are using it with a stalk quite happily, so it seems unlikely there is a hardware fault that has gone unnoticed until now.
The same dual opamp that is used for the stalk voltage buffering is used for the signal-strength amplification, and that works fine. So it's unlikely to be only one half of the LM258/KA258 that has died, although it's not impossible. Also, the PIC code, while it could become corrupted by power surges, is obviously mostly intact as the tuner works as a tuner. I can't see that it's very likely that only part of the code has died, so neatly confining it's effects to the stalk

It might be a construction error, I suppose. Can you check whether the diode D2 definitely has the end with the black band connected to +5v, and the other end connected to one end of R4? Also, check that R4 is really a 1K resistor, R2 is 22K, and R1 is 4.7K. If any of these were wrong you would get funny voltages at the PIC pin 2, which would produce erroneous results.
The stalk interface is a rather crudely generated analog voltage, from a resistor ladder in the stalk, and the PIC is simply reading a calibrated set of voltage ranges. If the source voltage is wrong, or the voltage produced by the stalk is out, you'll get odd/no effects.