Assumption 1: "The HD800's are regarded as one of the top headphones in terms of detail"
True, though sometimes all it takes is a frequency-specific emphasis to make it seem that way.
Assumption 2: "looking at the graph above, the SE535's and SE310's which both utilize balanced armature technology respond to the 50hz square wave response faster than the HD800's"
No and I don't see how you are inferring that from the graph. If you mean that there's a difference on the x-axis between the samples, that's only because it would be really annoying and useless to have to line up each headphone's data just right. There are marked differences in the magnitude of the leading edges and the geometry of their decay, but those shouldn't really be boiled down as a faster/slower thing.
Conclusion: "This then means those IEM's are capable of producing a more detailed sound than the HD800's, because they can respond to the amplifier's signal faster."
No, this would be wrong even if the second assumption were true.
Instead of picking small bits of data to arbitrate competing preconceptions, ask yourself, what made you come to think of them in the first place?