ars technica quote multiple food scientists at various universities , it wasn't a random ars technica writer. The problem with the way the CBC lab tests is that it used a DNA test to say what percent that chicken/soy etc. Problem is no one uses a DNA tests to calculate percent of content, they only use it to detect that the meat kind of meat. Example usage is detecting what kind of fish that is actually being served in a restaurant. Accurate test methods have a way of separating the meat into its different parts, ie dye methods or mechanical separating (example centrifuge) or spectrum analysis. DNA testing basically has too small of sampling to be accurate, ie you cannot do a large scale dna test of entire piece of meat, only very tiny portion that easily can be contaminated. Example they could sample a 100 different samples of individual piece of meat, but the problem is most likely even if half of those samples contained 1% soy, the test would come out with 50% soy content. Which isn't correct.