Also, science is persistently incomplete, and actually making decisions (or advice) requires making assumptions (often, neutral ones, but that can turn out to be quite wrong) about what is in the unfilled gaps. The advice to avoid peanuts was because it was clear that severe peanut allergies existed, it was clear that they affected a small fraction of children, and it was clear than when they affected very young children, those children weren't able to let people know what was going on as well as older children and adults to enable timely intervention.

There wasn't much information one way or the other on what avoidance did as far impacting development of allergies, and with the evidence available, delaying exposure seemed prudent.