This is something I have been thinking about for awhile.
My theory challenges that idea. It says that while in isolation one rubber may seem to fit a certain stroke better, the fact that you have different rubbers on each side adds to the computational complexity your brain's sensorimotor systems have to do, and therefore contributes to making more errors. I would add that using the the same rubber on each may cause strokes on each side to become more similar.
While I don't necessarily disagree with your conclusion, I do disagree with this premise. It has been my realization over the years that there is a strong overlay between pattern recognition and high speed reactionary performances. Human beings are remarkable at recognizing patterns. Be it faces, silhouettes, color gradients, It would seem we have evolved in in a way to recognize well.. better. Why is it that human beings can perform calculations such as projecting trajectories, reading spins, etc exceedingly fast? Because we don't and it only appears so! We are habitual creatures, who get better at things via iterative processes (such as drills, drills and more drills and, experience), as such getting good at sports like baseball or ping pong for us is about long term pattern recognition (we recognize arcs not because we are projecting the velocites/accelerations forward in time, rather because we've seen it before!). The difference between the experienced pro and the first timer is just that, the first timer is trying to crunch numbers, where as the pro has already done that math.
So you are probably wondering how does this tie in to my counter argument for the said premise. Because there is no added complexity! Getting used to a different backhand rubber than a forehand rubber is the same as the same rubber for both hands. In both cases you are recognizing, absorbing and internalizing two separate patterns, one for each hand. In the beginning, when we are building our tables in our memory, and we are relying on calculations to predict cause/effect then perhaps. But once it becomes habitual, it doesn't matter because in either cases you are establishing two different habits, one for your forehand and one for your backhand.
As proof of what I am saying, and something that clarifies the base case for your argument (I do agree, added complexity is real), consider the following cases - Kenta Metsudaira's tomohawk serve, left handed players, Zhang Jike's flicks, all these (and more) have one thing in common. Unknown variables. We struggle against left handed players or pips out players because we practice more against the opposite. When KM hit the scene, a lot of the pros (including the spin maestro Ma Lin) struggled, because he/they jhadn;t dealt with it before. I may be wrong, but didn't pioneer Zhang Jike the heavy reliance on flick returns? All these were new factors that pulled players (at the time) out of their routines and had them rely on calculations resulting complexity complication you mentioned.
My point is that while yes, there is such a thing as sensorimotor overloading (for lack of better words), the result is not from different rubbers, but rather from switching/twiddling rubbers. As long as external variables are limited, we can reduce the problem down to a memory that we can pull from our table (memory bank), but with new variables are introduced, we are back to stage one - fumbling due to sensorimotor complexity.
Also, one good thing about high speed sports, the reduction in time can also reduce the influence of external variables resulting in lesser environmental fluctuations, and actually make things simpler!