I am not sure that I agree that "the compressive force of the parts being assembled serves only to approximate the o-ring to the sealing surfaces, but any compressive force generated on the o-ring by assembly is not in the same axis as the sealing force at the o-ring generated by a pressure differential." I assume that without the vacuum system my housing is sealed with minimum pressure delta applied i.e. it will not leak in half an inch of water or if I apply a small overpressure into the housing (like heating it up in the sun) it will not leak even when the Delta P is in the opposite direction.
We design medical devices that really need to stay sealed and my engineering colleagues suggest to me that sealing, especially sealing with o-rings is not simple and that modeling the whole sealing process is extremely complex.
Yes it can be done, I wrote a short article that is on the Optical Ocean Sales blog site about it. Of course you can do it, using relatively bright LED lighting but you can't capture action and will have little DOF control.
I am quite sure that in TTL mode, the S2000 will ignore the magnet, it is only looked at by the strobe when in manual mode, so you can do both. I do occasionally use my EM-1 (with the same two strobes) in fill mode if I am shooting something that is moving and I have only one shot to take it. That being said, the frame rate is really slow 1/64 works for most of what I shoot.