- Develop 4 rolls in 1L of fresh, stock D76
- Develop next 16 rolls mixing the saved used-once developer with fresh stock at 2:1 ratio and using +30% development time
This is still just self-replenishment (for which D-76 isn't suitable, else we'd have been informed of it when Kodak and then Ilford dropped the replenishers for D-76 and ID-11). And again, it's a wastefully high replenishment rate; you get less film (20 rolls) than with 1+1 used one-shot.
I feel like everyone is over thinking this problem. OP can use his developer, then fill the used developer into his empty bottle. For the next stage he can just decant 50% of the used developer and mix in with 50% of his stock solution. After developing with this, then just discard the whole thing. The cycle then repeats where he can pour in the full stock developer and after developing the film, store it in the bottle, and then decant 50% and add 50% of stock buffer for the next development.
This and Old Gregg's method are both more complicated than proper replenishment, as well as requiring keeping track of whether you're using fresh stock or partially used developer.
By the way, I don't know why I prefer D-76 to ID-11, but I do. I feel more confidence.
I'd suggest starting by testing one against the other. Replenishment really is the most likely way to do what you want to do -- that is, have one development time all the time, and always get the same negative condition. Ideally, you'd use a
developer starter each time you mix fresh stock solution, to start in a "seasoned" state (there's probably a formula for D-76 starter around somewhere, and the same forumula would work for ID-11 -- once mixed, they're the same). That said, a lot of folks have used D-76 and ID-11 with replenishment since the 1950s without feeling the need for a starter.
The one big down side of replenishment is limited working solution life for traditional low-rate replenishment -- that is, over a long enough working life, the tank solution will build up enough byproducts to start to wander in activity -- you'll get an increase or decrease in contrast, changes in grain, bromide drag, or other problems. This usually happens when you've exceeded the recommended limit of equal amount of replenisher to the volume of the working solution (likely by quite a bit; manufacturer recommendations are generally very conservative, being aimed a commercial processing). The exception to this is high-rate self-replenishment, in developers that can use this method. The only one of those that's easy to get now is Xtol, though at least one of its knock-offs (LegacyPro EcoPro) has been user-tested to support this usage as well -- but with Xtol in self-replenishment, you can literally keep your working solution going forever (if you can avoid "trade concerns" and related product shortages).