Ima give you an analogy that isn't strictly the same but still illustrates a plausible mechanism or principle for benefits of IR.

Traffic. Lots of cars, all the time, lots of collisions. Solution, put permanent obstacles to slow down traffic, fewer collisions. Or so we believe, because we believe the primary cause of collisions is high speed. The fact is it's the speed difference between the two vehicles, but that's besides the point here. The point is, the more obstacles, the longer they are there, the more time any vehicle stays on the road, the longer this traffic lasts, the higher the probability of collisions between two vehicles, worse result.

The actual solution is to remove any and all obstacles, to allow traffic to flow freely and faster. This means for example to add parking space to reduce the time it takes to park, to reduce the time of this obstacle. Widen roads, smaller cars, overpass for pedestrians, etc.

Assuming we have the same total amount of cars, total time for traffic is going to be less. Normally, traffic is made up of all kinds of vehicles, including the cleaning crew, emergency vehicles and so forth. In a situation where the cause of the emergency is traffic itself, the emergency vehicles contribute to this same traffic, thus increasing the likelihood of more collisions, and then some more. Of course I'm not saying we let people bleed out in the streets just to flush traffic. The gist is that if there's no other vehicle on the roads, any single vehicle will be that much quicker on those roads.

So for our physiology, how this works. Chronic vs acute. Several smaller meals vs one big meal. Time of injury vs time of repairs. Bear in mind the point isn't to allow greater injury, but to give more time for repairs between injuries. It's a recurring event. In turn, more time for repairs means greater capacity to handle injury, or more specifically lesser effect from same level of subsequent injury, because the system is now stronger. If the system is always injured, never enough time to repair, the next injury will add to the first, making it worse and worse.

Now if we also reduce the amplitude of the recurring injury - by going LC - the system will always have more than enough time to repair injuries from the recurring events and from other potential events like infections for example. Repair may not be complete the first day, but the surplus - like the constant smaller injury that adds to itself - will add up each time.

I don't know if that's how it actually works, but it makes sense to me. Never mind the "we're adapted to whatever" argument, this is just a plausible principle.