If you don't see any patterns in your data, yet day-to-day fluctuations persist, he is reacting to something you aren't tracking. Look elsewhere.
As John's symptoms grew worse I became desperate. I began to doubt my earlier conclusions as I searched for another explanation. I asked the Feingold organization for help, and they suggested the <Failsafe diet>, a much stricter subset of the Feingold diet that avoids salicylates and amines, as well as additives. Amines are very subtle - you'll never figure them out without help. You have to worry about how long the steak has been in the freezer, and how soon the fish was frozen after it was caught, and how long the chicken simmered in the pot. If you haven't determined the exact cause of your food-related illness, and you have already eliminated artificial additives, the Failsafe program is certainly worth a try.
We ordered the book Fed Up, by Sue Dengate, and put John on the Failsafe diet, which is much more restrictive than the diet that had brought some success in January. Surely he would do well on Failsafe. But he got steadily worse. By mid April he was entirely unmanageable. Apparently amines and salicylates weren't the answer, or at least not the whole answer.
NFF: The failsafe diet is very high in carbohydrates, as it promotes rice, pasta, white potato, pears, and "all the sugar you want". John could not handle all these carbohydrates. His micro-colony thrived, and dumped their toxins into his blood stream. This completely swamped the benefits, if any, of fewer amines and salicylates.