I use C, slow mostly; A weighting is for lower noise levels so isn't as accurate, B weighting is actually what SHOULD be used at 80-90 dB levels (but can only be found on $$$$ meters :=(( so I use C - slow, because I want more the AVERAGE level at lower freq's in the room, and it takes some time for a room to respond at lower freqs.
I've mixed a few loud rock gigs (years ago) and between that and playing live quite a bit, I think that's where part of my hearing loss happened - that's why I've been being a lot more careful the last 10 years or so (be sure to close the barn door after the horse gets away...)
it's also kinda critical to get the SLM as close to your ears as possible, at least for periodic checks; X, Y, and Z axis locations will ALL change the results at lower freq's due to room modes, SBIR, etc. -
But the main thing is consistency of practices I think - especially when "pre-mastering" songs for a collection; if you want the songs to sound like they're part of an album, they need to be checked under the same conditions... Steve