OK here is how I did. First off, I didn't feel like listening to a test tone at 85dB tonight, so I set my monitors to 74dBSPL, placing the RS meter at my listening position.
I have an Apex 220, but I figured let's just use my KSM141, set to omni, instead. This mic is not flat in omni, it has a bass boost and a presence peak. But that's OK since I am using a test tone. If you were using pink noise, you'd need a flat mic. I put the mic in the same spot as the RS meter.
Second, to make things as bad as possible, I used my ART Digital MPA. Now the ART is a pre that for $400 I'm digging. However, for testing, it has the minus of have no fewer than 4 separate gain controls, and on top of that, I swapped the stock 12AX7 for a vintage 12AT7, which has lower gain. One big plus is a nice VU meter, which saves me the step of using the computer as a meter and having to figure out the A/D conversion, which ART unhelpfully does not publish in their manual (through experimentation, I determined that 0 VU = -12dBFS, but that is not important here).
OK, we're off!
The KSM141 is -37dBV @ 94dBSPL, so that should give me -57dBV @ 74dBSPL.
However, I needed +65dB on the ART to get to 0 VU (+2dBV), that would imply that the mic was producing -63dBV, and thus seeing 68dBSPL.
So I was off 6 dB, probably mostly because the 12AT7 is a lower gain tube than the 12AX7, so I wasn't really getting +65dB. Unfortunately, with my current interim setup, it would be really hard for me to calibrate the gain settings on the ART right now because I don't have an easy way of routing the converter out to the ART in
I am pretty convinced that anybody should be able to beat my 6dB error