Dear Merle,
After re-reading the attachment (all1) which I posted earlier, I realized that I had completely missed the most relevant section to yr OP since it came right at the end !. Perhaps you have already scanned the whole document more thoroughly than myself ? 
This last chunk (eg pgs 209 et seq) probably typifies, within some occasionally rather heavy maths, the "justification" for some of the elsewhere found opinions of "no risk" from soy contaminated wheat flour as linked in my recent post.
The text (implicitly) addresses my previously stated confusion as to why material which would seemingly qualify for precautionary Vital labeling is also categorizable as presenting no significant risk in terms of probable rates of allergic incidents. Unfortunately, as far as I can see, the explanation given is somewhat circular. It suggests that some flaws must exist in the theoretical risk assessment since significant allergic-incident rates are in fact predicted but comparable rates are apparently not seen in practice. Speculation over the nature of such flaws is illustrated in the following 2 text extracts. Similar comments appear in the presentation of both the peanut and soya case studies included. It's all rather confusing (to me). Perhaps some other reader here will extract a clearer interpretation.
It is necessary to read the details of the document to fully appreciate the twists and turns.
Soy protein was found at concentrations up to 236 ppm in wheat flour, a dose of 7.1 mg soy protein per 30 g serving. No published soy challenges have reported an objective allergic reaction from doses at or below 7.1 mg soy protein. Despite this observation, the quantitative risk assessment indicates that the user risk from soy commingling with wheat flour is rather substantial predicting 2850 reactions per day among soy-allergic consumers in the U.S. alone. Since no published reports exist of reactions that might be attributable to soy comingling with wheat flour, clinical allergists are either overlooking all of these cases (unlikely if 2850 occur per day in the U.S.) or the quantitative risk assessment is overstating the actual risk. A thorough examination of the inputs to the quantitative risk assessment suggests that are several factors that could result in an overestimation
. …………………..>>
However, the level of predicted risk (23.4 ± 4.9 per 100 soy-allergic user eating occasions) occurring if wheat flour contained the maximum level of soy contamination (10%) allowed by USDA grain standards is 100-fold higher. Perhaps consideration should be given to lowering these allowable levels of soy commingling to assure protection of soy-allergic consumers.
I later noticed that the above work is about to be shortly published with more details added.The preliminary abstract displays a rather surprisingly "no-need-to-worry" context reminiscent of my previously quoted links. The "get-out" from above mentioned discrepancy is ascribed to -
However, the predicted reactions occur at exposure levels below the lowest eliciting dose observed to provoke objective reactions in clinical oral soy challenges. Given this low level of predicted risk and the lack of evidence for allergic reactions among soy-allergic consumers to wheat-based products, the avoidance of wheat-based products by soy-allergic consumers does not appear to be necessary.
I suspect this is another way of saying that the "present" theory has "encountered" some practical limitations which take precedence.
The potential implication to Vital labelling decisions seems to not yet be considered.
Regardless, the content does indicate that the 10% tolerance mentioned earlier is applicable as a USDA criterion for (gross) wheat / soy cross-contamination. But the abstract does not mention anything about the "100-fold higher" mentioned above.
Some of the opinions inevitably involve a decision as to what is considered an "acceptable" incident rate, or to quote the unusual phraseology in Juliane's file, how to handle the "exquisitely" sensitive soy-allergic members of the population. Statistical logic unfortunately cannot avoid appearing quite "cold-blooded" at times.
Rgds / Charles.C
Edited by Charles.C, 26 November 2013 - 03:31 AM.
opinions rearranged