No, that's not how it works at all. You're still mixing up the general mathematical concept of ratio deco with specific implementations thereof. There is nothing inherent in ratio deco which shifts some time deeper.
I think the source of your confusion is failure to understand that there are two totally separate curves at work here. Ratio deco planning is divided into two phases, each with its own curve.
In phase 1 you calculate the total deco time based on the depth and bottom time. For a given set of gasses you can visualize this as a 3-D surface plot with the independent variables of depth and bottom time on the X and Y axes, and the dependent variable of total deco time on the Z axis. That surface has a complex curvature based on the chosen deco algorithm (ZHL-16C or whatever you prefer), but within any small region it can be approximated as a plane. So, you can establish simple linear relationships to quickly estimate total deco time to within a couple minutes of the actual deco algorithm.
In phase 2 you take the total deco time calculated in phase 1 and divide it up into discrete stops every 10 ft / 3 m. There is no particular requirement to follow a linear ascent rate (although that is an option). You can come up with whatever heuristic you like there to approximate the shape of the optimal ascent curve. Some divers follow something like a Fibonacci sequence (I am not recommending that specifically, just pointing out that there are multiple options on how to approach this phase).
The specific implementation of ratio deco covered in some GUE tech courses is optimized for pragmatism in terms of being easy to teach and minimizing the risk of dangerous errors in the field. They aren't trying to squeeze out maximum efficiency just for the sake of getting out of the water 2 minutes faster.
