The ideal spell to test something like this with would be a physical damage ability that didn't have a damage range. Thunder Clap, for instance, always does the same damage if you have the same AP every time you cast it, and so the only source of variation you'd get here is between a regular hit and a critical hit. An analogous Rogue ability in this regard would be Gouge. The other alternative is to get a weapon with a very low damage-range, and then use auto-attacks (or an ability that deals damage based on the damage of your weapon) to do your tests.
Confirming that the bug with Battle Stance still exists should be very easy. Equip gear such that you're sitting at around 91-92% Arp. Keep one of the sockets in your gear empty (this will be for the Arp gem you'll put in it later). While in Battle Stance, take note of Thunder Clap's damage each time. Cast it while facing away from the dummy, and make sure you dispel all your buffs before each Thunder Clap such that you always have the same AP. After getting a consistent reading of its crit and non-crit damage after 5-6 Thunder Claps, put a Fractured Bloodstone (12 Arp) gem into that empty gem socket you had. Observing any increase at all would confirm the bug, since if Battle Stance worked correctly, you would already be Arp-capped with 90%+ Arp, and thus not observe any further damage increases from going beyond that.
Pinpointing the exact source of the bug and how it works is a bit more complicated, in which you'd need figures for the boss' armor, its armor constant on Warmane (it used to be incorrect before the core swap, probably still is now), as well as your own Arp. There was a guide on these forums where formulas for most of that were posted.