In general, developing more gives more contrast, developing less gives less contrast. Now, when you throw dilution into the mix as a variable, "developing more" becomes a little slippery.
My experience (with Parodinal, which is a homebrew developer that uses the same times and dilutions as commercial Rodinal) is that agitation also has a significant effect on contrast, which allows playing off time and agitation against dilution. My preferred process with Parodinal, after I'd been making and using it for a year or so, was 1:50 dilution, 40% more time than common recommendations for the film in use in that dilution of Rodinal, but agitation only every third minute. This gives a slight speed increase (rather than the slight loss common for Rodinal), because it allows developing more while producing normal contrast. There's also a slight increase in perceived sharpness, because local exhaustion tends to increase contrast across sharp edges. Finally, agitation every third minute is enough to avoid bromide drag that can result from loner rests (semi-stand or stand development).
So, bottom line, what I recommend for Rodinal and its derivatives is maximum "standard" dilution (usually 1:50), extended development, reduced agitation, no change in EI, but alter development time by the usual percentages (20% per stop? Best to test this for yourself) to control contrast.
.EDU Ultra 100, Parodinal 1:50 (11 minutes, agitation every 3rd minute), Wirgin Auta 4.5.