The only things that were tweaked in the patches were drive, bass, mid, treble, (and, where applicable, presence, depth and MV) and speaker LF Resonance (to match the speaker used). To "match the real deal" we set the controls the same as the amp. That was the whole point of the exercise. No advanced parameters were used.
The IRs were captured using the Axe-Fx. Nothing special was done. I think we used a mix of an R121 and an SM57 but it might have been a KSM313 instead of the R121 (I personally prefer the KSM313 but can't remember if I got my way
). We placed the mics on the speaker, moved them around a bit and then mixed them at the console. The preamp was the usual API that we always use.
The one thing different about the IR capture is that we mixed at the console. For some reason this can sometimes result in really good IRs. The infamous "Factory 103" IR was mixed at the console. IIRC that was at least four mics that we mixed at the console. Mathematically you should be able to get the same results by mixing individual IRs in Cab-Lab but I've never really experimented with it much. I think part of it is the slight phase difference that gets lost because the individual IRs are minimum-phase transformed (but you should be able to replicate this by adding a little delay to one or more IRs).