I agree that he nailed it.  The problem is that he proposed a solution that wasn't necessary.  This is like the CEO who says the product has to go through another full QA cycle, costing the company another $350,000 instead of letting the possibly buggy code out because the devs know that the average 200 bugs cost about $25,000 to fix.  We don't bother spending the extra $325,000 just to please a CEO.  Plus, if Atlas was mining and he stayed with 0.7, he'd be (only very slightly) ahead of the game.  The 0.7 miners, consensus has declared, are the winners of the recently completed blocks, not the 0.8 miners.  The blockchain might not yet agree, but we're the sentient ones.