D-23 is possibly the simplest film developer in the entire library. Just metol, sodium sulfite, and water. The sulfite is doing triple duty here -- preservative, accelerator, and silver solvent.
But the replenisher for it is DK-25R -- which includes sodium metaborate in addition to metol and sulfite.
D-23 works fine at pH 8 of a sulfite solution -- is processing film adding enough acidity to require a significantly higher pH replenisher to offset?
Just guessing here, but maybe developing film adds enough more bromide to the mix that the developer needs a bit more accelerant to mix to counteract the restraining properties of the bromide.
Maybe a replenisher that contains only the ingredients of the original developer can't correct for some drop in pH. Most replenisher formulas use the same ingredients as the original developer but with quantities adjusted, usually to increase alkalinity (in addition to restoring developing agents).
Both reasonable hypotheses. I doubt I'll run replenished D-23 again any time soon (replenished Xtol is cheap, too, and I like the negatives better), I was just reminded by stumbling across sodium metaborate for sale at Freestyle.