The developing agents would be phenidone and glycin, which I don't believe appear like this. These crystals looks exactly like sodium sulfite to me. Potassium sulfite is much smaller and powdery crystals and plus potassium sulfite dropping out of solution seems unlikely with how soluble it is. Sodium sulfate also appears more fluffy than this. The solution was mixed without heating, so I'm unsure how it would become super saturated and then suddenly not be able to dissolve more. Sulfite and sulfate are nearly equivalent solubility for both sodium and potassium versions, but of course are greatly different between the sodium and potassium versions. This is why my best guess is that somehow sodium sulfite was formed. I understand equilibrium and that things are just ions in solution, but here seems complex given that both sodium and potassium ions are present, as well as glycol which metaborate is soluble in but not sulfite and I know glycol causes sodium sulfite to be less soluble (but unsure about potassium sulfite). I also know strongly alkali solutions also can cause sodium sulfite to fall out of solution.
I have no formal education in chemistry, so any explanation you can give would be helpful. It is not necessary to create such a strongly concentrated a developer, but I believe it contributes to the shelf life of the solution.
The supersaturation idea was a just a thought. I don't think it is highly likely, but perhaps something to keep in mind.
Warming up a solution to get more substance dissolved is the obvious case. I was thinking that maybe you may have managed to get a slight supersaturation. A border case. You are adding some components as solution. Also you are mixing a system of two solvents, water and glycol. Who knows what it'll do initially and then later, when it reaches equilibrium.
You seem to have a pretty good handle on chemistry, so I didn't want to get too detailed regarding the equilibrium of solution. No point in preaching to the choir. Essentially I was just trying to rationalise what you were already thinking that happened. You don't convert K2SO3 to Na2SO3 in solution, because there is no such thing in solution. But the dissolved ions can crystallise. And a Sulfite ion can crystallise with either a Sodium or a Potassium ion swimming around. Both will happen. That is more or less the 'reaction' that you would have in mind:
2Na^+(aq) + SO3^2-(aq) <--> Na2SO3(s).
2K^+(aq) + SO3^2-(aq) <--> K2SO3(s).
This process is running in both directions, as do all reactions. Formally at least.
The solubility of the solid is the constant that determines the equlibrium of this system. I.e. which side you will end up having in your bottle.
Equilibrium means that the process never stops. As an example: a saturated solution of NaCl with crystalline NaCl at the bottom looks static. But the solid keeps dissolving, slowly, while ions are being deposited at the same speed. Like a full nightclub. Five people come out, so five people are allowed in.
The same process will happen with the Potassium ion and the sulfite. For K2SO3 the solubility is much higher, so the dissolution will be dominant and you'll have essentially no solid. For the Sodium salt we are not sure. That is the question here. What we need to know is the solubility of Sodiumsulfite in a solution of Potassiumsulfite at the concentration in your developer. There are a few more other components, all of which will have some influence. As you say, this is a very complex system. But this is something that you could test if you have both chemicals. This will only be an approximation because in your developer the Sodium ions come from another salt and the sulfite concentration remains constant. You could also add NaCl to the K2SO3 solution and see what happens.
I am not working as a chemist - I ended up in IT, but I was a chemist, originally. That is why I like to do a little self mixing, occasionally, and doing darkroom work in general.
So I can only delve into the theory of what could be happening. I don't have that much practical experience on this particular problem. I did read about solubility issues in the Film Dev Cookbook and did wonder about it. DD-X was the prominent example, I think. Some sodium will obviously be tolerated, and, as I have tried to explain above, I believe this will work as long as the respective (much smaller) amounts of the (formal) sodium salts are still soluble in the overall solution.
It might be more informative to see if it is soluble in alcohol to confirm it is not a product of glycin or phenidone. And try glycol to see if it is metaborate. That would leave only sulfite, sulfate, or carbonate finally and sodium or potassium could be tested by how much is soluble in a fixed amount of water. And sulfite/sulfate could be determined from carbonate by testing pH. The real difficult one to test is sulfite vs sulfate, since the drying process would likely oxidize sulfite, as I don't have a vacuum chamber or any other clean way of drying the powder otherwise. I'd also have to make a fairly large batch of developer and purposefully let this precipitate issue happen, since the amount of powder, even if relatively significant, is still such a small amount from my test batch that it'd be difficult to work with.
An alternative that I may try is using potassium metaborate (check a previous thread I posted in for synthesis details) which would pretty quickly determine if it's an issue with using potassium vs sodium salts, though of course I expect most people would not want to mess with the synthesis process, even if relatively simple.
If you had very fine needles it could have been a hint toward organic compounds. In your case it could be anything, but a salt will be most likely. I also wouldn't waste too much effort on determining if it is sulfite or sulfate, at least initially. You said that the solubility of sulfite/sulfate is very similar. Hence, the solution for the solubility problem shouldn't be affected. The question of oxidation will be another problem you may have to care about, or not. It is one of the jobs of the sulfite to be an oxygen scavenger, after all. There should be enough there to continue maintaining its other functions.
