Rodinal has always used the free base in the developer itself, right from it's initial formulation, Dr Andresen's contemporaries were well aware of this. It certainly appears they didn't want the Chlorides present in the developer.
If I was making a developer like the old type Rodinal, Calbe R09 or similar, on a commercial scale I think I'd mature it with a very slight excess of Hydroxide, and then use a small addition of Potassium Metabisulphite to remove the excess hydroxide. You could back titrate a sample to determine the exact amount to add, once there's no excess hydroxide the free base begins to precipitate, but the pH will also probably drop at the same time. It's hard to know without trying it.
Judging from Patents on p-Aminophenol manufacture it appears there must also have been issues with controlling the purity during manufacture, so possibly other tests were conne before finalmadjustments & bottling.
Commercial alternative p-Aminophenol developers were available by the early 1900's Johnson's Azol was an early alternative to Rodinal, again using the free base. Ilford introduced it's version Certinal in 1909
Kodak's approach was to use p-Aminophenol Oxalate, they sold this as Kodelon and presumably it's the developing agent in "Kodilon" which was their highly concentrated liquid developer, and competitor to Rodinal. Kodak also sold p-Aminophenol hydrochloride as Kodelon. I have only ever seen references to "Kodilon" in UK Kodak Formulary as a commercially available liquid para-aminopheol type developer.
It appears when p-Aminophenol hyrochloride is used in combination with a second developing agent and excess carbonate there's less of an issue with the Hydrochloride, and it was used in other Agfa, Kodak, Ilford etc developers.
Lowe used "Gradol" which was the Edwal company name for p-Aminophenol Sulphate.
I deliberately didn't give the formula for J. Desalme's "Concentrated Paraminophenol Developer of the Rodinal Type" because there are some issues with the weights & conversions & concentrations used in the British translation of L.P.Clerc's book. The UK editor of the book George E. Brown, the Editor of The British Journal of Photography, and also the British Journal Photographic Almanac, died in 1934, and the book wasn't published until 1937 some considerable time after his death. (It is a large and comprehensive book). G.E. Brown was the Tangye Scholar for Chemistry 1892-3, at Birmingham University and would have had a full understanding of Photographic Chemistry.
However Brown also published the same formula in the British Journal Photographic Almanac.
Paramidophenol hydrochloride 68 g
Water, hot, 600 - 700 ml
Filter solution if necessary.
Sodium Sulphite 9.36 g
Sodium Carbonate (anhyd) 31.2 g
Water to 200 ml
Add B. to A. The paramidophenol is thrown down.
When mixture is cool filter off the deposit on cloth, and let the paste dry until its bulk is not more than 300ml.
Then, in a graduate, mix with it 100 ml of soda bisulphite lye 35° B. and add strong solution of caustic soda of 40° B. (about 50%) until the base is just dissolved. Water is then added to make 500 ml. The solution is diluted 20 to 30 times for use.
Sodium Metabisulphite (soda bisulphite) and Sodium Hydroxide (caustic soda( were commonly sold as a solutions.
The Sodium Metabisulphite 35° B, is approx a 28-30% solution and it is still commercially available in this form.
The Sodium Hydroxide 40° B, is closer to 38% than the 50% given.
The yield of p-Aminophenol free base is approx 50.97 if conversion 7 recovery is 100%.
The Avoirdupois & Metric conversions/proportions are correct, this formula uses about 10% of the metabisulphite advocated in other similar formula which allows a higher concentration of the developer.
I've only added it to the thread to show how easily the free base can be formed.