A+B developer dilution ratios.

Let's say you have a two part stock developer such as FX-2, and you decide to dilute the developer to, say, twice as dilute as normal. Under what circumstances would you want to change both the ratio of reducer(part A) to total volume and the ratio of accelerant(part B), vs changing only the amount of reducer(A) and using the normal ratio of accelerant(B) to total volume.

A similar scenario: if you want to use a highly dilute developer, say, Rodinal 1:300, and you mix it from scratch, couldn't you make a dedicated batch that uses a lower reducer to accelerant ratio in order to maintain buffering, in an attempt to maintain film speed?

If you change the ratio of B to A in the dilute solution you will not have FX-2 but something else. Calculating the expected developing time in this case would have to be trial and error. Most people would keep the same ratio.

If you are mixing a Rodinal type developer from scratch you cannot change the ratio of hydroxide to p-aminophenol and expect the developer to either have the correct activity or to keep. While the Rodinal recipe may look simple, the chemistry is actually rather complicated.