i wonder about the Sigma's ability to relate TTL info to the flash if you were in fact in TTL mode on your flash. i am not positive about how TTL info is gathered and communicated but because of this i would do 2 additional tests:
1. shoot a canon 85mm (if you can get your hands on one) with the sigma 85mm just as you did with the 70-200 and see if the same problem occurs
2. shoot the same test again in manual flash mode with the sigma 85mm and canon 70-200mm (this should yield the exact same results...if it doesnt something is terribly wrong)
also, have you repeated the original test and gotten similar results? if not i would do it again just to see if it was a fluke. sometimes there are connection issues with my canon flashes on canon bodies that cause the flash to go out of TTL mode and default to manual flash power at full power. if i take the flash off and put it back on the camera it will then recognize TTL communication again.
This is certainly a repeatable event. I would say I switched back and forth between lenses 5-6 times and had the same situation happen each time.
I took a brief gap at lunch to try setting the *flash* (originally said "camera") in manual mode. I can't be certain that this would have fixed the issue last night, as the lighting in my living room is now drastically different. But it did result in similar exposures, lens-to-lens, with the same settings. This seems to indicate to me that the flash is firing with less power when the zoom is attached compared to when the prime is attached when in E-TTL mode.
I did note that my settings in the camera were "evaluative" metering for flash, as opposed to average. Not sure if this will make a substantial difference.
Thanks all for weighing in. I will continue to experiment with different settings. If need be, the flash can just be operated in manual mode