Hi HD_Fatboy
An non-pleasing miracle "after counting down the right bonus: Then faulty counting down the left bonus". At the moment I have not an idea worth writing about.
I would like to do some tests - I believe to read in / from Your post-8: You do tests with the playfield lifted - and You manually trigger / simulate "ball rolls into the right Eject-Hole". You write "manually trigger ... then activate a second time" --- the way the pin is constructed: YOU MUST keep the switches actuated / closed - ALL THE TIME the pin is busy giving points / stepping down the bonus-ladder. Do NOT tip-on and take Your finger away --- the pin NEEDS the switches closed all the time.

You wrote "There are TWO set of switches" - Yes, see the JPG: "encircled green" is the switch to make the right Bonus-Relay to pull-in (it then stays pulling as the bonus is stepped down). When You have bonus: The situation is "as my drawn brown lines, a bit below in the JPG" - when stepping from 1 bonus down to zero bonus: The Eject-Relay must pull-in.
The second switch on the right Eject-Hole is encircled blue --- it is for: The pin MUST immediately step-up to one bonus to end the feature "ball has landed in the right Eject-Hole".

Please do testing again - park Your finger (hihi - I try to evade a joke) in the Eject hole - have the finger there until the kicker kicks Your finger - THEN take the finger away. And then write a new description "this is happening". Greetings Rolf

0Grand-Prix-Work-33 (resized).jpg

Hi HD_Fatboy
thanks for doing the test. With "parking the finger to close the switches permanent" we made little progress as (only) the weird behavior of "two times closing the switch needed" has vanished --- but basically we are back on the problem / description "Your post-1".
An important information is Your answer to "Did ever Your pin was AAA: fully functioning - then the fault started to happen --- or did You BBB buy "the pin with this fault" ?
When You say AAA: We can exclude - when You say BBB: We must consider that an preowner made some strange / creative wiring --- (((the schematics only shows "original wiring")))

The ipdb-manual http://mirror2.ipdb.org/files/1072/Williams_1976_Grand_Prix_Manual.pdf tells me on page-12 (ori-10): left and right bonus-unit are mounted on the playfield, left and right bonus-relay are mounted on the playfield, left and right Eject Kickers are mounted on the playfield --- but the Eject-RELAY is mounted on the bottom panel in the cabinet.

You wrote "LEFT (Eject-Hole) feature fully works" - the design is "counting down the left bonus - then the Eject-RELAY pulls in AND the LEFT bonus unit is stepped up one step - then the ball gets kicked out" --- question: Does Your LEFT feature ends with "stepping the LEFT bonus unit up one step - then kicking out the ball" ?
In post-15 You write "testing right Eject-Hole-feature: The Kick-out-Solenoid fires (as it faulty starts on the left))) - question: Is the RIGHT Bonus-Unit stepped up one step when the kickout-coil fires (((the schematics tells me: Every time BOTH kickout coils fire as they are connected (due to parallel wiring) )))
Am I allowed to say "a foreign person looks at the pin - his description: The ball rolls into the right Eject-Hole - the right bonus is counted / stepped down - the ball is kicked out, the right bonus is stepped up one bonus - ball is moving on on the playfield --- strange, the left Eject-Hole-feature then happens" (((If we do not find the cause of the fault: I am tempted to try to do an work-around - suppressing the faulty "and then the left Eject-Hole-feature happens")))

Please write about:
Does the bonus steps up one bonus (from zero to one bonus) when the feature ends on the left side ? And on the right side ?
Does the LEFT Bonus-RELAY pulls-in and stays pulling at the very moment the faulty "left side" starts happening ? Greetings Rolf

Hi HD_Fatboy
I hope for a miracle - we fix on the pin what is not fully functioning - and then (miracle) the fault with the bonus dissapears. (((And the other faults You sure want to fix at a time - lets do it now.)))
I choose "left Eject-Hole-feature" --- the Eject-Relay shall not pull-in when there are some bonus on the left bonus-ladder - and the ball rolls into left Eject-Hole. On the JPG (taken from the ipdb-manual) we see the left bonus unit drawn in position "Zero bonus on the unit" --- adding bonus makes (step, step, step) the wiper-arms to turn counterclockwise.
"my green line" shows the connection "Zero Bonus" --- "my red line" shows the connection / position seven bonus. Then the ball rolls into the left Eject-Relay - the bonus unit is rhythmically stepped clockwise, stepped down --- see: "for many steps there is connection "encircled-13 to encircled-15" --- and then on the last step "ONE-bonus down to ZERO-bonus": NOW the connection is made "encircled-13 to encircled-14" and NOW the Eject-Relay is allowed / must pull-in (((not: before that last step down))).

Please have a look at the unit - wires soldered-on --- BR-BLU and BLU-BR and W-B --- somewhere a short ? wipers not centered on the rivets ? wires running from "at the out-rim soldered-on - running into the unit - soldered-on at the rivets ?

While I wait for Your answer: I study / write on "Bonus-Unit should been stepped-up from Zero to ONE bonus at the very end of the feature". Greetings Rolf

0Grand-Prix-Work-34 (resized).jpg

Hi HD_Fatboy
great - You found the fault. I gladly accept the miracle. Maybe (maybe) it was like I show in the JPG: See "my red line" extending to the connecting line "wire in the pin-color-BR-BLU-3". When the right bonus is counted down reaching zero position: Connection (my brown line) is made to make the Eject-Relay to pull-in. FAULTY along the "my red line" the left Bonus-Relay is made to pull-in ...
In normal operation the fault is slumbering because open are the two "encircled rosa/pink switches" and the "encircled green switch".

On the top of the topic - above post-1, to the right: There is an icon to mark topic "as solved" --- I just happen to like to see these marks - please mark topic "as solved". Greetings Rolf

0Grand-Prix-Work-35 (resized).jpg