Just perusing some old books on developing and some of them don't have any minimum developer amounts for various formulas.
Any formulas or rules of thumb for this problem?
Required development capacity by film surface is an elastic parameter... in theory the developer dose should be able to develop all developable silver in the medium to obtain film DMax in every spot in the frame.
In practice usually only a fraction of the silver is developed so usually only a fraction of the developer is required, but dosing developer in excess holds chem concentration stable so development is more consistent.
So IMHO it depends, if we produce thin negatives then we require less developer, for bullet proof negatives style way more developer is required or usually recommended.
What we can test is development capability, with samples of a certain film, lights open, we may fully develop the samples with different doses of developer and measuring the density reached by the samples, in that way we may compare developing capability of different developers, this does not explain all effects, of course, but this tells
what relative safety margin we have with different developers at different doses.
I guess that tabular grain films require less developer than classic cubic films for the same results, as grains are flat a higher density may be reached with less developed silver, so with less developer consumption.
Kodak with Xtol recommends starting with at least 100ml of developer per 80sqi (a 120 roll, or 135-36, or four 4x5 sheets), this is a conservative amount... when I made a mistake using 2/3 of that I found no difference, but if we have specially dense negatives (say a lot of surface in Z-IX) then it would be interesting to see the effect.