One common plank of the freewill defense is the contention that, to be meaningful, our choices must have predictable consequences. That entails a stable environment. And that means God can't jump in to save our bacon. The laws of nature are the price we pay for responsible decision-making.
This argument was popularized by C. S. Lewis. Although it's common to the freewill defense, a Calvinist can also incorporate the same principle into his overall theodicy. Making choices based on predictable consequences is certainly consonant with compatibilism.
So I think the stable environment theodicy has some merit, but it's inconsistent. Although nature is predictable in some respects, nature is unpredictable in other respects, and some natural evils are among the least predictable features of nature.
We've gotten pretty good at forecasting hurricanes. (Of course, that's useless to our pre-scientific forebears.) But we can't predict earthquakes, mudslides, tsunamis, droughts (and drought-related famine).
If a tsunami occurs, we maybe able to predict the trajectory and arrival time. However, that's generally useless because you can't evacuate port cities in a few hours.
We can't forecast tornadoes. If they occur, we can track then, but the lead-time is often down to minutes. There really isn't time for advance warning. (And, or course, our prescientific forbears couldn't even track tornadoes.)
We have a far better understanding of disease transmission, so we can now avoid or reduce some epidemics and pandemics, viz. rat control, draining or spraying malarial ponds and swamps. (Again, that knowhow wasn't available to our prescientific ancestors.)
There are genetic diseases we can now predict, but not prevent. So the patient feels doomed by a dreadful diagnosis and fateful prognosis.
It's hard to protect against venomous snakes. If you see them in plain sight, you can avoid them or kill them with a rock or a stick.
But take Indian farmers who are bitten by cobras when they harvest rice patties. That's predictable in the coarse-grained sense that there's an appreciable risk. But it's unpredictable in the fine-grained sense that you don't know where not to step until it's too late.
Likewise, it's well-nigh impossible to snakeproof a house, especially huts, shacks, and shanties. These are porous. Lots of ways for snakes to get inside.
If it's a nocturnal venomous snake, you may be bitten if you step on it in the dark. Or you may be bitten when it crawls into bed to snuggle up against that nice warm body.
So there's often no way to protect against kraits, mambas, cobras, Russell's vipers, &c. You can't take adequate precautions, or anticipate where they will strike.
The stable environment theodicy is a poor fit with what the natural world is actually like. At best, it intersects with a part of nature.