Very interesting indeed.
I mixed five liters of XTOL a month ago, and as I had read from hrst's experiments earlier, I decided to leave one liter to the lab and put the rest to refrigerator to wait for use. I just don't shoot enough film to use five liters of developer diluted 1+1 in six months. I just checked the ones in the fridge, and they have same kind of fluffy goo as in Ihmemies's photo. I'm going tonight to the lab, I'll check the remaining developer there for the flakes, if I remember.
I also mixed the developer with local (Espoo) tap water, so it's more than likely that the flakes are some kind of organic products. Mold, fungus, or something. It is quite clear to me that oxidation, or storage in soda bottles, have nothing to do with this stuff.
I actually doubt that the flakey stuff hasn't got that much to do with your storage life problem, but it just happens to occur the same time. (Though, it also could be that the flake-causing-organic-creatures are consuming some vital developer component.) I will get back with my results when I have developed something in the refrigerated XTOL.