I believe you misunderstood my point. My rebuttal to the "cushion" argument isn't about whether or not it is reliant on the healer, its that the cushion exists already on all the tank invulns, its just whether or not the healer decided to use it. Yes, using the pre-tank invuln time to deal damage takes skill and coordination, and yes that is about player skill, but the fundamental argument is that it is still there.
In another thread someone talked about Holmgang not having a huge benefit in o5,6,7 because the tank busters aren't even reducing them to 1 HP. So I looked at a random run of o7s and what was holmgang used on and it appears that the player actually didn't need healing for up to 30 seconds before their use of holmgang. And in this case the healers actually healed up all the remaining damage before the tank buster and shielded for the tankbuster and then healed the same damage a second time, which is exactly the point you made. What could have happened:
30 seconds before the tank buster, healers could have ignored the tank with maybe a regen to keep them from getting too uncomfortably low (even this would have been unnecessary), getting down to 10% or 20% HP, let the tank invuln, giving up their remaining HP, and then the tank could have been oGCD healed (using the oGCDs which were previously used to heal in anticipation of the tank buster), shielded, and regened to get back up wasting relatively few resources and buying the healers multiple damage GCDs or time to watch the bleeds on the dps.
Would the 10 seconds of "cushion" on the dark knight invuln have changed anything significant in that example? No, maybe it would be up earlier later on, but thats the real value of the "cushion", timing. It takes skill to use, and it takes coordination, but that doesn't make it bad. My argument is having the cushion is not the huge benefit it was being made out to be, not that it is bad because it relies on healer skill.