I'm wondering now if pH can be used to vary the volume of D23R per roll in order to maintain consistent D23 pH or is the increasing pH actually necessary (and as designed) to keep the other ingredients active?
To an extent. The effects of the use and keeping around of developer are multiple:
* pH shift due to interaction of the developer with the film itself
* oxidization of the developing agents by exposure to aerial oxygen and through the development action itself
* leaching of halides from the developed film into the developer through the development action
If you start with bare D23 and then proceed towards some kind of replenishment system, it's evident that you can only compensate in part for these factors. After all, halides will build up to a certain level, which supposedly should stabilize with an appropriate rate of replenishment. I don't know what the makeup of the D23R replenisher is; apparently it's a higher pH than D23, which I guess should compensate for the halide buildup: the halides act as a restrainer, while the higher pH will increase activity, so to an extent, these effects could balance out.
I say 'could balance out', because getting this exactly right seems virtually impossible to me. I can see how you might achieve a sort of stable process over time, but I don't think it'll ever be as stable as just using the developer one shot. Given the modest cost of formulas like D23, I always considered that to be their main appeal. If there's some image-wise benefit to replenishment, I'd be inclined to just mix a different developer that employs some restrainer and uses a higher pH right from the get-go, and use that one shot. You'll quickly find yourself re-inventing the wheel, of course.
Here's a hint that your replenishment process might be doing what you think it does:
I noticed that D23 pH increased at a rate closely matching the volume of D23R used eventually reaching a similar pH as the D23R. The effect on development is consistent with increasing pH and I learned along time ago to shave development time about 1 minute/500ml of D23R based on film tests.
This suggests that your final developer is something in-between D23 and D23R, but closer to the latter. Maybe that's intentional, but I've never quite understood what the appeal is in working with developer that provides inconsistent result for a significant number of rolls, only to (hopefully) somehow balance out at the end. YMMV.
As to trying to track pH to determine appropriate replenishment rates: by measuring pH, you're only assessing one side of the balance. You'd ideally have to determine the halide/restrainer content as well. I don't know of any feasible way to do this in a home darkroom environment.
PS: pH meters tend to be iffy when used at relatively high pH's. Make sure to take proper care for yours; store the probe in the appropriate solution and regularly calibrate your meter if only to verify that you're still getting consistent readings. A drift of 0.2pH seems minor and can easily occur, but is big enough to make the reading entirely useless (or worse - it'll lead you out into the woods).