As assumption should be tested with the negation test. Ask yourself, if you negate option E and plug it into the original question, will it negate the argument?

In other words, even if "most consumers who drink Mango Paradise will NOT eventually experience serious cardiac problems" (that is, some consumers may, but not most) does that still not make a valid case for banning the product? Should you only ban a product if MOST consumers will develop medical problems (and not ban it if its affecting only some)? The answer would be, even if SOME people will develop cardiac problems from the product, it should still be banned. Hence option E fails the negation test. Therefore thats not the answer.

Lets apply the same test for C.

If "any drink that contains at least 2.5% of Ephedra is NOT harmful to consumer health" and this particular drink contains only 2.5% (which is not harmful), then the conclusion (that the product should be banned) becomes invalid. Hence C is the answer.

When author says unconditionally that if 2.5% is the quotient available in drink , its gonna be harmful. Option C is the best... if he had put some other catch e.g. harmful to males/females/ area population etc.. may be we could explore other options.