In other words despite many Jewish grandmothers who advocate its use it really doesn't do anything.
Now I would dispute that, purely from the medical field in which I work. There is a most definite placebo effect it exerts a psychological benefit to the
patient photographer rather than for any
physiological practical effect if administered with sufficient conviction.
I suspect the Jewish grandmothers knew that, if delivered in a convincing way the benefit can be clearly demonstrated. This is one of the primary reasons why in clinical trials those in patient contact are blinded to which arm of the trial their patient is on, active or placebo (more usually these days the current gold standard treatment as giving nothing is rather ethically frowned on now) to avoid influence.
I was involved in a placebo trial recently and we had three patients removed from the trial early for unacceptable side effects reported by the patients and considered so serious they were withdrawn by their physicians, you know what's coming, yes all three, when the code was broken, were on placebo.
This is why, way up there, I suggested a parallel film exposed under identical conditions in the usual developer to compare, with only the experimental film and high expectations it is likely the result would be biased positively especially as the evaluation was to be subjective.
Bottom line, I am as unconvinced as you by pre-washing/soaking but I do use it where it is recommended by workers who have exhaustively tested their work and I am using their process, I have seen no work that demonstrates detrimental effects and as you say both Ilford and Kodak are neutral would tend to confirm that.
And finally, before we have the inevitable angels on a pin comment from somewhere, the
density of angels dancing on the point of a pin is dependent on the assumed mass of the angels, with a maximum number of 8.6766*10exp49 angels at the critical angel mass (3.8807*10exp-34 kg).
Quantum Gravity Treatment of the Angel Density Problem
by Anders Sandberg
SANS/NADA, Royal Institute of Technology, Stockholm, Sweden