I cleaned up an original Blackface Princeton Reverb a couple of years ago and wrestled with the same questions. The amp shocked people, so a 3-prong cord was first on my list. The filter cap for the bias circuit was toast and the power tubes were biased really hot. The owner wanted adjustable bias anyway, so I pulled the entire bias board and installed a new one. Easy enough to restore (other than solder joints). Scratchy tone pots lead me to leaky caps in the tone stack - those got replaced once I narrowed it down to 2 caps. One cathode bypass cap was obviously leaking goo and the paper cover rotated freely, so I replaced that too. After those changes, the amp sounded fantastic so I left everything else alone.
What does the owner want? A noodler/collector has completely different needs from a studio guitarist.
Unless the owner tells you to replace everything that might be a problem down the road, I'd only take out what was absolutely necessary. That old rectifier tube is worth some money if it still does its job. Leave it alone.
Unless the power resistors have drifted so much that voltages are way off in the power supply, I'd leave those in there too.
Screen resistors - the original didn't have any so if this amp already does, someone has been inside there before you. The amp I worked on was untouched, but I did add screen grid resistors.
I mapped out voltages throughout the amp, all of the resistor values and the caps that I could measure in circuit. Most of the resistors had drifted up in value, some more than others, but none were completely out of bounds to me. (A 300K plate resistor for a 12AX7 clearly qualifies as "out of bounds" for me.) Other than the leaky caps, the rest were reasonable too. Of course, they'd generally drifted to lower values. Sure, some voltages were "off" but none more than 20% after I was done.
Any chance you have time to map voltages, both on tubes and on the power rail? I'd be really interested in plate resistor and cathode resistor values too.
Just in case you're interested, I've attached the bias board layout I used. I did have to drill one hole in the chassis - with the owner's approval first! Had to tweak a resistor value to get the right voltage range (18K net) and added a 100uf cap in parallel with the second filter cap (attached to bias pot) to clean up the bias voltage ripple.
Have fun in there!
Chip