If you look at the schematic, the entire metering section, JFET, 9V battery is floating and isolated from the rest of the multimeter (divider strings, shunts, rotary switch etc.)
Only Points G, A, E between them.
It seems to be around 2.7MEG worth of leakage yet R25 is 5MEG so what's with that. But changing the Polarity switch and getting 3.3MEG for a different leakage resistance of near 5MEG now seems baffling as well.
OP what type of multimeter are you using, to know its Ohms voltage and test current etc. if it is throwing things off.
Point "G" should be high resistance unless it's pushing enough to bias the JFET junction or (reverse) turn on zener D1. Check R25 is 5MEG.
Ensure there is no lithium grease used on the rotary switch, that is a no-no. Lithium is a salt and can be conductive lube. I would clean the PC board with Q-tips and IPA, get rid of old flux, soot, carbon etc.
I think it's hard to tell if a component's value is low or it's contamination.