On behalf of a colleague:

With the test being open boxes and in most equipment about 50% of the volume taken up with components I wonder if the IE out in front is more or less than calculated? Does the reduced volume direct more energy out front or do the components absorb some of it?

It's less. Sorry I can't find the paper on MCC arc flash testing right now. But basically this is a "radiosity" problem (google this). Every surface absorbs and reflects some thermal radiation. Since nothing is a perfect reflector (aka white body), some energy is "lost" in the process...absorbed by the components. It doesn't matter if it's MCC's though or something else. If you look through EPRI's database of papers and read the ones on their arc flash testing, they do mostly "actual equipment" tests. The results are lower than what IEEE 1584 calculations suggest, sometimes dramatically lower. It really isn't a matter of volume...it's not a "hot air" argument. For the most part it's all heat transfer by radiation, with ONE exception.

The exception is that if there is a substantial "barrier" such as the top of a circuit breaker or busbars clamped by a piece of insulated board such as a heavy support or bracing between compartments, occasionally the arc sits in one spot instead of propagating further down the bus bars and the hot gases pool and then jet outwards. When this occurs the incident energy can be 200-300% higher than the normal "open box" case. However arc rated rainwear actually protects better against this than the more common open weave fabrics. IEEE 1584-2018 now addresses the increased incident energy but not the differences in PPE performance where IEEE 1584-2002 didn't address it at all.