I agree with the gist of aegert's comments, but also understand where Moke is coming from. As tapers, I think we have to ask ourselves what our vision or 'listening goal' is in capturing the audio of a show and the techniques we use to do that.
I record SBD+AUD 4track or multi-mic mixes a lot more often these days than stereo AUD.
With my 2 track stereo AUD recordings, the only post work I'll do is to peak normalize (add gain to -0.2dBFS) and maybe a high-pass filter if called for. My 'goal' with 2 track AUD is to capture the room as it was that night. I'll only adjust the EQ further using a ParaEQ in the rare circumstances that it sounded so awful in the room, or my recording location, that the EQ adjustment is called for to make it listenable on mediocre playback systems. With multiple AUD pairs in the same location (on the same stand, etc.) being mixed to stereo, I'll rarely add more than a limiter or some light (2:1) compression to reduce just the highest transients that tend to be increased by the summing effect of mixing the sources.
For anything other than ambient stereo AUD, however, I will apply some 'light' post-production as I feel it's called for since I consider my 'goal' a little differently. With multi-source recordings that are not full multitracks (i.e., stage + PA AUD, SBD + stage + PA, SBD + AUD, etc.), I see the goal as more like using the tools at hand to make a shorthand type of multitrack recording where I'm trying to represent a 'sweet spot/front row' type sound that is more than the component stereo sources may convey. I want the crispness of the direct sound if I have the SBD feed, yet I still want it to feel like you're in a room with other people and to represent what it was like to be at the show in a way that the SBD tracks alone would not be able to do. With these recordings, I typically add 'light' (2:1 - 2.5:1) compression with the threshold set a little under the average peak level. I follow that with a ParaEQ, typically adding a high-pass filter and possibly using a notch or shelfing filter to reduce a bit of some frequency/frequency range as my ears - and maybe RTA viewing - indicate is needed. Depending on the peaks of the strongest transients, I may add a limiter before I peak normalize to -0.2dBFS. I like the average rms levels to be in the -20dB to -16dB range, thus leaving space to somewhat accurately respresent the dynamic range of the show as it may have been heard in the audience.
*edits for spelling