The 60 / 30VAC issue is due to how you are measuring the incoming voltages.
The 12 and 14VAC is measured with respect to ground (black lead on ground).
The 60VAC does not have a center tap so it must be measured between the two input pins.

Regarding high output voltages:
The high 60/42VDC could be due to no load on the outputs. Stick a 600 ohm, 10W resistor between pins 4 and 8 of J3 and see if it drops down to normal.
600ohm, 10W resistor is fairly common resistor found on Bally rectifier boards. This will stick a 100mA load on the output which is higher than normal but still good load to test with.... and it will get hot!

And the two offset voltages -- 0.08 for both. That one does appear to be the problem. This provides an offset to the display filament voltages. Could be they went out on the board or you can have a short somewhere. Trace wiring back to the transformer for your two low voltage filament voltages and look for loose wire, short, etc. The RD most probably generates those two the same way everybody else does... a simple pair of diodes/resistor for the 4V and a zener diode/resistor for the 8V. Check those on the RD board.

Unless... The RD still has the old isolated grounds design (I certainly hope not!). If it does, make sure you are measuring these two voltages using the correct ground. Measure those offset voltages with respect to the ground on J2 instead of J3.

Shokky diode?!?!

Anyways -- looks like two grounds. Copied the original Gottlieb design. Gottlieb liked to isolate grounds and use a star configuration for tying the grounds together in the bottom of the machine... and we all know how well Gottlieb grounding works. Isolating the two types of grounds on the power supply was unnecessary but at this particular point it only caused more confusion than problems.

Remeasure the 4V and 8V but use the ground from J2.

Nope, I wouldn't go to 100 ohms.
You have 16.5V coming in with an 8.2V zener diode. A 100 ohm resistor would result in a current draw of 83mA and would be dissipating 0.7 watts (with a half watt resistor).

The 470 ohm resistor would be 17mA and roughly 0.15 watts. You don't need a lot of current here - enough to break over the knee current for the zener diode which would be just a few mA. Any more than that and you are dissipating excessive resistor and diode power. The 470 ohm resistor here is fine.
I wouldn't go any higher than that, though. A higher resistor has less current. Too high of value and the 8V load current can overpower the voltage drop across the resistor and can actually drop the voltage lower than the zener voltage.

Could be a shorted zener. Or it could be a shorted 8V output on J3. I would first give a quick glance to the wiring between the power supply and the transformer.

Typo.
J3... not J1.

Quick question -- is this with or without the connector installed on J3?
Without connector installed - definitely bad zener.
With connector installed - still could be external short.

The 74V on the input is actually a bit low.... but the output of a regulator can't be higher than the input so something is being measured wrong.
Measure those voltages one more time - black lead of meter to pin 5 of J3.
Red lead of meter to:
Banded end of CR7 to measure input voltage.
Banded end of CR12 to measure output voltage.

As mister Cactus Jack said -- if you managed to kill the 8V zener which is on J3 then something else on J3 may have also gone awry at the same time.

I thought the design flaw of the RD Gottlieb board was fixed when he changed U1 from an LM317AHVT to a TL783CKC (along with associated circuitry). You might have one of the older boards.
The Ni-wumph / RD issue isn't really an RD issue... nor is it really a Ni-wumph issue. Neither one has a defective design where this problem is concerned... the two boards just don't like each other. Has to do with switch on speed of the RD power supply versus reset timing of the Ni-Wumph board. To fix this - the RD board could add a "soft start" capability but I don't think his 5V regulator supports that. Or the Ni-Wumph board could revise their reset to a solid state power up reset. Either remedy would require PCB changes which aren't cheap.

Yes, there is an increase but you're looking at the wrong place - it is the 60VAC that is being rectified, not the 74VDC (which is already rectified and filtered).

The input voltage is 60VAC. After the full wave bridge rectification (measured across the input cap) -- you will have roughly 60 x 1.414 or approx 84V for your peak voltage (unloaded). This is where the approximate 90VDC comes from, an approximate value because the 60VAC input voltage is an approximation. This is DC voltage at this point, not AC voltage. Make sure you are measuring with your meter set to DC voltage and not AC. If set to AC voltage, you are measuring your ripple voltage - a different animal.
This 90VDC should be going *into* the regulator but you list 74VDC as the input voltage. The 74 is below normal.
Coming out of the regulator, you listed 95V. This is not a step up type voltage regulator and an increase in voltage is not possible on his output. That is why you have either a voltage measurement problem or you are feeding high voltage back to the regulator outputs within the wiring.

Those two photos were more than all the verbage above that post.
Like Mr Cactus Jack said --> shorted regulator. End of story, look no further.

Input of 74VAC -- you should get (74 x 1.414) - (2 x 0.7) = 103.2 volts on your input. Rather high but not the issue as the System 1 regulator should be able to handle this. And you see nearly the full voltage on the output (102.7) -- the regulator isn't doing anything ... it's shorted.

Look at U1 -- what is the part number of the existing regulator?
If it is the LM317AHV - it's a known problem with that design. The LM317AHV can withstand an absolute maximum of 60 volts between input and output voltage. During normal operation - you have 102 volts on the input with an output of 60 volts --> a 42 volt differential and still within spec.
BUT if you were to ever short the output and during power-up period, you would have that 103 volts on the input and 0 volts on the output. Part would be running at 175% of absolute maximum ratings - that does mean and nasty things to these regulators. The National Semiconductor field application engineer (hopefully, they all found new employment by now) - actually said the power-up period would probably kill more of these than the shorted output.

If it's the LM317AHV, see if Jim has a resistor mod to change that to the more robust Texas Instruments TL783. This part can withstand a differential up to 125V. Same pinout (if I remember correctly) but requires different resistor values.