One of the reasons I hate this fight is because I play a main tank shaman and would love to get the update for my conjuror. We only ever do this fight if we have a new applicant that needs the update for mythical, and that's it. The problem with that is no one wants to play an alt healer for this fight and none of the non-MT group healers want to be in that group for the fight. I have absolutely no issues dealing with the mechanics of the fight, and have only screwed up once (where I simply didn't get my group noxious cure quickly enough back when we were learning the fight). At this point I'll end up just paying one of the farm guilds to let my conjuror die and get the flag that way (though I have no issues legitimately playing him for the update, if anyone would have him along for it!)
Another reason is that due to us only doing this fight when new people need a mythical update, we inevitably have people who have never done the fight before, never played the character they play as their main on the fight or are a bit rusty on the mechanics. I rarely screw up during the fight, but I cringe when someone makes a tiny mistake (doesn't get noxious cured in time, flubs sprint and doesn't toggle it off while they are feared, someone doesn't click the cube or statues) or the statues bug out and go active.
It is a great training encounter and I think it was well done, but I do agree they went way overboard with the amount of failure conditions. Is it really necessary to have one person be dedicated to clicking statues the entire time while someone is ready to hit the cube at the correct time along with everything else that is built into this fight? There's no skill or challenge involved in making someone click those statues or even click the orb at the right time. If those two facets were removed completely, some bugginess would be removed from the encounter altogether. If those mechanics must remain in some way as part of the lore, just make it so the cube needs to be clicked before anyone is able to hurt him or something.