This latest discussion started with a very general intuition that as people are asked to perform more and more work they become increasingly fatigued and their physical and mental performance suffers.
That's an intuition most of us have just based on our own general observations and personal experience. And it's natural to apply that intuition to what's going on in football as well: maybe there's a measurable performance advantage gained by a defence where its offense holds onto the ball longer, thereby sustaining the D's performance by allowing it to perform less work.
Recognize however that so far all we have is the intuition. It's strongly held but it really tells us nothing. We don't even know for sure that it does apply to what we're seeing on the football field: maybe an elite level, heavily conditioned athlete with proper hydration and nutrition can play 20-60 snaps (depending on rotations) over the course of 3 hours and perform more or less as well on the last snap as on the first.
Percy has presented at least preliminary evidence to suggest this "fatigue/freshness" factor is not operative. I think before the discussion goes much further it becomes incumbent upon those arguing for the existence of the fatigue/freshness factor to show why we should expect it to be operative.
His own arguments suggest the approach that might be taken. One could start, as he did, by searching for evidence of a correlation between offensive TOP and defensive performance (as measured by, say, points per drive). Percy's data sample was obviously small but if one were to do some regression analysis using 10-15 years of data (with offensive TOP ranking as the independent variable and defensive points per drive ranking (or whatever performance metric one prefers) as the dependent variable) one would obviously get a better sense of whether offensive TOP correlates with defensive performance, beyond limiting opponent drives (using per drive defensive performance metrics effectively backs out the opponent drive-limiting effects of offensive TOP).
One might find one gets r-square values so low that they override one's basic intuition that performance suffers as more and more work is performed: the numbers might convince one that, for whatever reason, that intuition doesn't neatly apply to defensive performance on a football field (which is what Percy's very preliminary data suggests). On the other hand, one might evidence of a correlation in support of our intuition that is worth exploring.
But until one has such evidence I don't think it's fair to hold up our naked intuition that increased workload leads to poorer performance because of fatigue as something Percy's analysis fails to account for. He's done enough to show that, without further evidence, it's not even clear that there's a real performance advantage gained by a "fresh" defense for which he needs to account. Show him otherwise.
