"Rodinal" was originally the trade name for para-aminophenol hydrochloride, and Agfa's developer was thus called Rodinal developer - just like others were metol developer, hydrochinone developer, and so on. After a very short time the Rodinal name stuck to the particular mix made by Agfa.
There are many old published recipes for Rodinal developers. The only thing they have in common is the developing agent, "Rodinal". That may be the source of some of the confusion.
If the quantities of bisulfite and hydroxide had contained the same number of K/Na ions it would have made no difference at all. They don't, but I still think it makes no difference! The main reason for using potassium salts in concentrated developers is that these usually have higher solubility than the sodium equivalents. In this case it should all dissolve anyway, unless you use the sodium forms of both.