Yes, you're right about the pictures. Here are some:
EDIT: my pictures were too large, and didn't upload. Here they are now, though!
This is SN 4220786; this meter was purchased on ebay this year(2020), and from the chip date codes I think it was new in 1987 or so. This is the one that I already removed C19 from, but it had been installed with the negative post toward the display. You can see the "+" on the ckt board for C19. I also laid the old C19 on top of the buzzer.
Here are some photos of SN 5135102, this meter was inherited from my Dad, around 1999. From the chip dates, I think this was new in around 1990 or 1991. The photos are an overall shot of the board, and then a closeup of the C19 area. It looks pretty clear that the C19 negative is installed toward the display
This is my 3rd 8060A. It is SN 6520032, and I purchased this unit new in the mid 1990s. Chip date codes make me think it was produced in 1995 or so. This one also has C19 installed with negative terminal toward the display.
This is a shot from Mrmodemhead's blog, and appears to show C19 with the negative terminal away from the display:
http://mrmodemhead.com/blog/wp-content/gallery/f8060a_01/IBM8060A_013.JPGAnd finally, here is the schematic, which shows the "+" side of C19 tied to Pin 6 of U2. I confirmed that on the board itself, the terminal labeled "+" goes directly to U2 Pin 6.
So, what I can't understand is how 3 different meters, produced over a span of 8 or 9 years, could have this capacitor installed backwards, and still work just fine!
Unless the capacitor, in that location, does not see any DC voltage? I'm not knowledgeable enough in circuit analysis to be able to tell...