That's a reasonable way of explaining, but I still disagree, and I'll give you may take on it. I'll also provide how I think it should be addressed with additional mechanics. The first is the AI recognizing RCI and acting to deal with it. The second is a series of intervention programs, wherein the AI spends its budget on improving a particular RCI.
Again, 3 systems. 1 backwater, 1 advanced, and 1 backwater + Hydra assisted. Again, let's assume they all start at 0, and then suffer a massive plague that drops medical by 100.
Backwater: Medical drops to -100. They have nothing to modify RCI, so they stay a desolate wasteland. A new government comes in and decides that the worst problem facing the planet is Medical. They then invest in the "Train New Doctors and establish Medical Protocols" intervention program. This slowly increases their Medical. Eventually they are able to get back to 0.
Advanced. Medical drops to -100. They have advanced hospitals (buildings) and advanced training (technology) that automatically improve RCI. Their Medical increases as these passive bonuses work. They decide that they aren't dealing with the situation fast enough, so they also kick in the "Train New Doctors and establish Medical Protocols" intervention. However, because of their technology level, their version of this intervention is higher, so their RCI goes up even faster than the Backwaters.
Backwater + Hydra: Medical drops to -100. They have nothing to modify RCI, so they stay a desolate wasteland. However, the Hydra Player takes mercy on them and starts a Medical dispatch. A new government comes in and decides the Hydra is taking care of things well enough, and they should spend their budget on new statues for the palace. The Hydra player is able to get their Medical back to 0 and then goes off to stop a war. The Backwater is happy with their brand new statues and then decides to repaint all of the houses on the planet purple because the green paint they used last year is cursed.
Okay, so then the question is should RCI be a unbounded float or a percentage?
I believe that RCI should be unbounded because things can always get worse or better. I could see it being on a log scale, but that's an extra level of complexity that is beyond the immediate situation. I don't think the percentage is the way to go because it forces an unnatural cap. If we don't bind it between some x & y, then it is the same as not being a percentage and so we can skip the percentage and have the exact same outcomes.
And should it decay naturally to 0 over time?
Again, I don't think so. My first impulse is that it is adding another level of complexity too soon. Getting a basic functional system where buildings, technologies, events, and player actions consistently and sensible affect the RCI should be performed first. Then afterwards if additional complexity to RCI is necessary should decay be considered.
When would decay make sense?
Let's do some examples. Let's say you have a hospital it adds +1 Medical per year. You have it for ten years (RCI is up to 10), but then it is destroyed. It makes sense that RCI should degrade in that situation as your Medical Training dies off. However, consider a second example. You accidentally unleash a planet wide radioactive cloud that drops the environment to -1000. How quickly should your planet recover from this? Should it recover at the same speed as the loss of the hospital loses Medical? If we have a simple multiplicative reduction or a percentage reduction, then the radioactive cloud would reduce more quickly than your people forget about medicine. That seems wrong. A reduction in radioactivity would take centuries, not just a few years. The other big problem is the time scale. It really seems that the basic unit of time should be years rather than months. Given that our races can jump from unmanned space probes to time travel in a few years. As you can see, it is complicated. Hence my suggestion to make a simple working system, and then try to be fancy about it.
