My personal suspicion is that the levels varied considerably depending on time and place and the number of calls active at any given time.
You may want to contact the Telephone Museum in Seattle, which has an operating #5 crossbar. If they have any sort of traffic simulator hooked up to it, it should be possible to measure SPL under various simulated loads. It might be an interesting exercise.
Note that during much of the era when crossbar switches were in use, exposure-related deafness was not very well understood. Not surprisingly, a lot of the early research on environmental hearing loss was done by Bell Labs. --scott
"C'est un Nagra. C'est suisse, et tres, tres precis."
[TELECOM Digest Editor's Note: I know that in the early/middle 1960's the Chicago-Wabash central office -- in fact all of Chicago -- was either crossbar or stepping switch and you could hear it a block away on a hot summer night walking down the street, mainly because they did not have air conditioning in those days and all the windows would be open wide. When it started to rain, someone would go around and close all the windows. Quite deafening. PAT]