OK, I'll try to be yet more precise. There are actual white jokers that light up near the top of the playfield; a rollover lane and a standup target. Then there's the white insert down by the kickout hole that's worth 1000 when lit.
So when you say "white joker lit", are you literally referring to the lit white jokers (rollover and standup target) at the top of the playfield, or do you really mean that the white insert near the kickout hole is lit? (Only the pair of white jokers, *or* the white insert should be lit, as those lights are controlled by a make/break switch on relay 2B.)
And even more specifically, the white joker relay is relay 2B in the sequence bank, which is the middle relay in the picture you have above. As shown in the picture, that relay is in the "untripped" position. In that position, the two white jokers (rollover and standup target) should both be lit, and the white insert near the kickout hole should not be lit. While 2B is in this untripped position (along with 1B and 3B), the kickout hole only scores 1000 points. Is that correct?
Then when that relay 2B is tripped (and only 2B), instead of getting 1000 for the kickout hole plus 1000 for having achieved the white joker targets (which should now both be not lit), it's getting 3000 total. Is that correct?
And finally, if you block the switches on 2B while it's tripped, do you still get the 3000 points? Or do you only get the 1000 points? This is what I'm mostly trying to understand:
2B not tripped = 1000 points (from kickout hole)
2B tripped, no switches blocked = 3000 points (the problem)
2B tripped, switches blocked = ? points
If it's still giving 3000 points with 2B tripped and all switches blocked, yet only gives 1000 points just by not having 2B tripped, then that suggests maybe the switch blocking is incomplete, or some other physical short is created from tripping 2B.
Another thing to try would be to not trip 2B, but instead jumper closed only the switch that completes the 1000 point scoring path on 2B (SL+GR, OR+GR per the schematic). That *should* cause it to score 1000 + 1000 without having tripped 2B.
The purpose of these experiments is to get a different behavior, to help further isolate where the problem is.