This is interesting theoretically as well as practically.
So practically, my view is that if you're using DNP, the objective cannot be to build new muscle tissue (as ATP is going to be dramatically reduced: thus the energy state of the cell directs energy metabolism to catabolic processes).
As such, just use a - here we go again - minimal effective dose. So that means, don't go under 100 mg daily (no real justification; even at 100 - 200 mg daily, DNP is less efficacious as an insulin sensitizing/antidiabetic agent than aspirin in humans; and at 100 mg, expect a modest ~10% RMR increase [less in heavy bodybuilders]).
Certainly, a keto diet can rationally serve this objective: in an individual with a low b.f.%, with the use of MCTs, et cetera, keto can be used to really reduce those difficult to reduce fat depots like the lower back in men & the hips/thighs in women. DNP will make this all the more attainable (and we can "get in and get out," staving off longer-term muscle catabolism).
Theoretically the reason that DNP does lower blood glucose concentrations (but remember, it's a minor effect in humans) is by uncoupling of oxidative phosphorylation ⇒ ↑↑O₂ consumption (reflects the increase in RMR) & ADP:ATP ratio, ↑lactate synthesis (the latter due to a compensatory ↑glycolysis due to low ATP, thereby ↑pyruvate [gycolysis end-product]) & ↓glucose (due to the ↑glycolysis): hence hyperlactacidemia is a serious concern with DNP use.
So yes: with DNP, keto makes sense (i.e., entering ketosis); anticatabolic agents that are selective in the intracellular compartment (androgens) make sense as well; the utility of rhGH here is complicated by its hyperglycemic and insulin resistance effects, but it certainly makes sense for longer-term adaptations to caloric restriction (i.e., ↓IGF-I): but again, the use of a burst or plateau busting tactic for short term "diet sprints" at already low body fat seems the most logical approach for DNP & keto.