While I had the case open, I checked the FG-801’s calibration:
Look at all those parts!
The raw filtered DC power supplies run a bit high and the output voltages & frequencies were off by a little, but not too much after all these years.
Page 11 of the instruction manual gives the setup and calibration adjustments (clicky for more dots):
Page 12 gives some values that should be true:
For whatever reason, the manual isn’t available on The InterWebs, so here it is for your amusement:
Fordham FG-801 Sweep Function Generator – Instruction Manual.pdf
The Smell of Molten Projects in the Morning 
That is some great vintage technology there. Look at all the 741s :-)
Reminds me of an EE Lab assignment from back in the day, where we were tasked with taking an Intersil 8038 function generator chip and working around some of its major quirks (might have been removing a dependency between duty cycle and frequency, IIRC). That chip had a lot of, uhh, surprises in its behavior.
I remember that for extra credit, we were to see if we could find and document any unpublicized discrepancy between something on the datasheet and the actual behavior of the chip. I randomly picked a parameter, not having any particular reason to suspect an issue, but trusting that the part would not let me down, and then tried to verify the datasheet values. No worries, the chip did indeed not match the spec. I think it was frequency deviation versus supply voltage — but that was long ago in the misty past :-)
I never worked for Intersil, but the 3 other places I worked for treated specs verrry carefully. Absolute maximum meant “if you exceed this, the chip will almost certainly fail spectacularly”. Guaranteed values were just that, within the temperature ranges specified (usually junction temp, and determining that could be a challenge). Typical values might or might not be accurate. If it wasn’t guaranteed, watch out.
At least in the ’70s, a fair number of companies didn’t publish guaranteed reliability numbers (in non mil-spec conditions–the DOD had a big say and a firm opinion for that class of parts). ’70s automotive parts were, er, interesting. We had one customer (auto company and vendor not named to protect the guilty) who insisted on putting the engine controller in the engine compartment. IIRC, they had a max temperature higher (140C) then mil spec (125C). We delivered working parts–at least when they left the plant.